This week I wanted to try something new. A Q&A from the phage community! Sometimes we get questions from labs, researchers, companies, about our thoughts or advice on a particular area of phage work.

Here’s a question I got from a researcher setting up a phage therapy center in their lab, and my (real, lightly edited for clarity) reply that I sent to her.

I thought I’d also share it here so all can benefit!

Question:

As you know, we are currently setting up a small-scale lab for personalized phage therapy production. I would be very grateful if you could share the specifications or recommendations for the equipment you used [during your time at Phage Australia], particularly the ultrafiltration (UF) system.

My answer:

Hi, thanks for asking! I love this question! The equipment I really like for phage production and quality control is:

For tangential flow filtration (TFF)

• AKTA flux S (see my blog post on it here)
• Required consumable: Midgee Hoop Hollow Fiber Cartridge 100kda (the cartridge that I used on the AKTA flux)

For column chromatography

• AKTA Pure 25 (FPLC system; same as many use for protein purification - basically the base model is all you need, but there are various components you can choose between depending on how flexible you want to be)
• Consumables:
  • OH-1 monolith hydrophobic interaction column from Sartorius BIA separations (Slovenia) (I follow this method for filamentous phages, haven't tried for lytic ones yet, but I think it might work); and I follow this method for lytic phages)
  • Endotrap columns (Lionex, Germany) (I use this instead of the OH-1 column if I ONLY want to remove endotoxin; otherwise OH-1 is ideal if you want to remove endotoxin and also host proteins, DNA, etc); for these to work you need to do TFF first; for the OH-1 column methods, you can do that straight from LB-containing lysate.

For verifying endotoxin levels at the end

• Nexgen endotoxin reader from Charles River - this is the best! And can buy FDA-approved cartridges. We just got one. (Regular price ~10k, we got a hefty discount, so I leapt at the chance; have to buy cartridges so each sample will cost around 40$ to test; therefore this reader is ideal for final product testing but not if you're doing lots of samples as part of R&D)

If you have limited budget:

• Instead of AKTA flux S, you can get a Sartorius Vivaflow (and secondhand or new peristaltic pump) (see Tiffany Luong’s awesome paper on this)
• Instead of AKTA pure, you can use octanol (see Phage on Tap paper) to remove endotoxin — we did this for some of our patient preps in Australia, worked well; just use dialysis to remove octanol after)
• Instead of Endotrap columns for AKTA pure, you can buy the gravity columns and not need the AKTA pure at all - this works well, just needs more time
• Instead of Nexgen endotoxin reader, can use Biomerieux Endozyme II kit or similar - more laborious than Nexgen and not FDA approved, need to validate yourself / can be annoying to set up, but doable
• Instead of TFF you can use Amicon spin columns (see Phage on Tap paper) and dialysis after you've got a purified prep

We are going through process of trying to get funding for these here at Stanford. That said, we are likely going to get Vivaflow cartridges and a peristaltic pump (1-3k online) instead of AKTA flux for now. And we already have an old AKTA Purifier (old model of the AKTA Pure) so we are using that for now - works great, just it may die and when it does, we will have no manufacturer support. We have the Nexgen reader now but also have the Endozyme II kit for when we have a lot of samples.

Hope this helps, happy to help if any of this is confusing! If you tell me your approximate budget I can suggest what you should choose among all those things.

Jessica

P.S. For those who want to dive deeper, here are the slides from a talk I gave on using TFF for phage purification at Evergreen 2025 (which just happened!) for the Phages for Global Health Phage Purification Workshop!

Send us your questions!

There you have it — our first Phage Community Q&A post. Let us know how you like this format so far!

And if you have your own question, or something you want advice on in the phage lab (or computational space), or a piece of equipment you want advice on before you buy, email us at capsid@phage.directory! We might not know, but we’ll see if we can get your question answered by someone in the phage field!

Phage Production and Manufacturing · Technical Aspects of Phage Research · Tools

What does it take to make phage therapy work in clinical practice in 2025? Who better to ask than the clinical lead at IPATH, the well-known phage therapy center run by UCSD in San Diego, which has set the example for phage therapy centers in so many states and countries since.

On this week’s episode of the Podovirus podcast, Joe and I were excited to sit down with Dr. Saima Aslam to understand how her approach has evolved since treating her first patient in 2017, and what her advice is for other phage clinicians.

The answer, it turns out, lies in a fundamental shift from "treating anything and everything" to strategic patient selection, as well as prioritizing close collaboration between phage researchers and clinicians. This way, trials can be designed that make sense for patients, clinicians and scientists.

One thing I really appreciated hearing was Saima's assessment of where the field stands now: "I feel like this coming decade, where we have enough knowledge to design good trials, I personally think is going to either make it, or if we have multiple failed trials, that's the end of it."

I agree with this sentiment, and am excited to see what these next few years hold for phage therapy.

Highlights:

• Patient selection evolution: How Saima's team moved from treating all-comers to targeted approaches, learning which infections respond best to phage therapy.
• The importance of collaboration: Why bringing together phage researchers and clinicians from the start leads to better clinical trial design with more realistic endpoints, so patients can be more easily recruited (they of course have lives to work around too!).
• Dosing insights: From treating phages like antibiotics every 8 hours, to understanding serum neutralization, biofilms, and the importance of local delivery for device infections, and adjusting dosing to fit that.
• Finding the sweet spot: Why recurrent UTIs in kidney transplant patients represent promising applications (even low hanging fruit, according to Saima) for proving phage efficacy.
• Infrastructure needs: The case for a centralized US phage repository and manufacturing center to eliminate the current 6-12 month delays that make compassionate use challenging for many cases.

A few snippets from the episode...

On the reality of complex cases:

"Even with the best phage, if we can't change the underlying disease, I'm not sure treating the infection will help the patient get what they want."

On the stakes for transplant patients:

"To have waited 10 years to get a kidney transplant and then you get recurrent UTIs and you lose it in two years is awful. It's just terrible... If you don't have recurrent UTIs, your kidney is good for the next 10 years."

On trial design challenges:

"Some of the companies I've seen start trials have done so because they know the phage piece really well, but they haven't really researched that clinical piece well. I think that's a big reason why some of these studies have failed."

On endpoints and timing:

"I think if we work collaboratively, if we can show it works in one thing, then we certainly will have a lot more money to figure out how it works in something else... We should pick sort of the lowest hanging fruit rather than the hardest ones to go after."

On working across networks:

"I literally send out emails to whoever can help a patient... Texas A&M, Baylor College of Medicine, Daria Van Tyne’s lab in Pittsburgh, Graham Hatfull... It used to be mass emails, but now they're more targeted because we know who really has phages that are ready to go."

On the current phage-making bottleneck:

"Initially you would think ‘maybe this is a patient I would treat with phage’, and you would get phage a month later. Now there's such a long backlog of cases that it takes more than six months, sometimes up to a year to get phage."

What's next:

Saima is leading an NIDDK-funded placebo-controlled trial for recurrent UTIs in kidney transplant patients. Her approach involves treating asymptomatic patients to target colonization rather than active infection, and the team just began enrolling this week!

Concluding thoughts

To me, this conversation crystallized why patient selection has become such a recurring theme across our phage therapy interviews. On one hand it's driven by the need to ration a scarce resource, and not get patients' hopes up unnecessarily, but it's also about using phages only where we know they can succeed, so we can build solid evidence and insight quickly, and use that to design trials that are actually doable. By showing we can do this, we can bring in the funding and gain the experience to properly test the next set of indications, and expand out from there.

Thanks so much Saima for your time sharing these lessons!

Listen to the full episode:

🎧 Spotify

📺 YouTube

Learn more:

• Saima's team is enrolling for their kidney transplant clinical trial
• IPATH at UC San Diego
• Our previous episode with Mayo Clinic's Dr. Gina Suh: another US physician’s perspective on patient selection and clinical experience

Clinical Trials and Case Studies · Phage Therapy Applications · Phage Therapy Access and Implementation

Hi everyone,

We found some interesting papers for this month’s Picks!

My biggest takeaway is that the more we learn about how phages, bacteria, and the immune system interact, the more we realize how intricate and complex the entire system actually is, and that there’s a lot more to uncover and understand.

~ Jan

Phage-induced protection against lethal bacterial reinfection

What is it about?

This study shows that phage therapy does more than just kill bacteria — it also trains the immune system to fight off future infections! Researchers used mice infected with a lethal dose of E. coli (a sepsis model) and treated them with a therapeutic phage called HP3. Not only did the phage cure the initial infection, but when they re-infected the same mice 4 weeks later (without giving them any more phage), an amazing 93% survived what should have been a lethal dose. The protection seems to come from phage lysis of the bacteria and releasing antigens that the immune system can learn to recognize, kind of like a personalized vaccine made from the patient's own infection.

Why I'm excited about it:

I think this is huge! We've always thought of phages as just bacterial killers, and thought about their immune responses as an annoyance if anything, but this research shows they might be functional as dual-action therapeutics — both treatment AND vaccination rolled into one. The protection was incredibly strong (reducing bacterial loads by a billion-fold!) and didn't require any additional phage treatment (and the original phages had long gone!). The fact that this worked with two completely different phages (though HP3 was best) suggests this might be a broader phenomenon we can harness — and that some phages might have this property more than others (and maybe we should pick those for future therapy).

~ Jess

Paper: https://www.pnas.org/doi/10.1073/pnas.2423286122

Xing, Y., Hernandez Santos, H. J., Qiu, L., Ritter, S. R., Zulk, J. J., Lahowetz, R., Patras, K. A., Terwilliger, A. L., & Maresso, A. W. (2025). Phage-induced protection against lethal bacterial reinfection. Proceedings of the National Academy of Sciences, 122(22), e2423286122.

Temperate phages enhance host fitness via RNA-guided flagellar remodeling

What is it about?

This study reveals how certain temperate phages (called Flagellin Remodeling prophages or FRφ) use RNA-guided systems to completely redesign their bacterial host's flagella. The researchers found that these phages produce proteins called TldR that work like molecular switches to turn off the host's normal flagellin, and replace it with the phage's own version. This flagellar redesign gives the bacteria extra properties: they swim better, hide from the immune system more effectively, and colonize the gut more successfully. Using cryo-electron microscopy, the team showed that the phage-modified flagella have completely different structures than normal ones, explaining why they work so much better.

Why I'm excited about it:

This paper actually blows my mind a bit, especially as I studied flagella and (Campylobacter) phage interactions in my PhD — I would have been so psyched to see this then! At the time it felt like it would be crazy to propose that phage were having a hand in what the flagella did… Also, the fact that these RNA-guided systems evolved independently multiple times suggests this is not a fluke and could be a broadly useful strategy. Very cool to think that phages could be helping their bacterial hosts become better at surviving in real-world environments like our guts. I wonder how this might lead to new therapeutic approaches where we could use phages to modify bacteria in beneficial ways.

~ Jess

Preprint: https://www.biorxiv.org/content/10.1101/2025.07.22.666180v1

Walker, M. W. G., Richard, E., Wiegand, T., Wang, J., Yang, Z., Casas-Ciniglio, A. A., Hoffmann, F. T., Shahnawaz, H., Gaudet, R. G., Arpaia, N., Fernández, I. S., & Sternberg, S. H. (2025). Temperate phages enhance host fitness via RNA-guided flagellar remodeling. bioRxiv.

Predicting clinical outcome of Escherichia coli O157:H7 infections using explainable Machine Learning

What is it about?

The authors in this paper collected a bunch of shiga-toxin producing E.coli (STEC) genomes from across the UK, and used two machine learning approaches to predict clinical outcomes of the STEC infections. Using both Random Forest and XGBoost, they were able to correctly predict which isolates were high risk, in what would otherwise have been low-risk lineages.

The researchers were then able to connect the predictions back to what genomic elements led to the predictions (which included phage-encoded elements, and others).

Why I'm excited about it:

I think papers like this show that you can use prediction models to see your phage and isolate bank in different ways, and find new patterns and connections. In this paper, the models found potentially novel intergenic features as highlighted by their “explainable” models (as in, the prediction tells you which features, e.g. k-mers in this case, might be a reason for its high prediction rate). So using methods like this is kind of like presenting key ring of the most viable keys (out of thousands), but you still have to check them. Some doors are already opened, like stx2a in this case, but others could open doors you didn’t even know existed!

~ Jan

Paper: https://www.medrxiv.org/content/medrxiv/early/2025/06/06/2025.06.05.25329036.full.pdf

Paganini, J. A., Khatun, S., McAteer, S., Cowley, L., Greig, D. R., Gally, D. L., Jenkins, C., & Dallman, T. J. (2025). Predicting clinical outcome of Escherichia coli O157:H7 infections using explainable machine learning. medRxiv.

Phage Biology and Research · TechBio · Phage Therapy Applications

Happy July!

This month, Jessica and I have been very much heads down in deep work, both in phage work and AI work! We’ve been adjusting things in our mail provider Mailchimp, and unfortunately a couple of newsletters got messed up for this past month.

A lot has changed since we started Capsid & Tail many years ago, and we’re considering making some more changes to the newsletter — please let us know what you’ve liked and disliked so far!

Meanwhile, for this month, here’s what we published:

You should sequence your phages

by Jessica Sacher

In this article, Jessica emphasizes the importance of sequencing your lab’s phages. If at any point you’re in doubt about what phages you’re working with — when you receive a batch from collaborators; when phages sit around for too long; heck you should be sequencing them ideally every time you propagate your phage. The end goal isn’t necessarily to sequence your phage: the goal is to at all times know what you’re working with.

Inside Canada’s first prosthetic joint phage treatment

by Jessica Sacher

This was a cool podcast interview Jessica did with Dr. Marisa Azad, a physician behind one of Canada’s first bone infection phage therapy cases (and it was a success!). Of note is the discussion of choosing an N-of-1 approach vs. a compassionate use route, and the pros and cons of each approach. Both choices have their own merits and their long-term benefits and consequences — and for Dr. Azad, she chose to lead a multi-center trial platform for phage therapy, across Canada. Please give this one a listen!

Producing phages consistently: then and now

Throwback: The phage that helped treat Dr. Marisa Azad’s patient came from Slovenia, and this is a good time to re-read how Frenk Smrekar and Barbara Hubad of JAFRAL help companies and university labs consistently produce safe phages. Worth a read!

~ Jan & Jessica

Technical Aspects of Phage Research · Clinical Trials and Case Studies · Phage Production and Manufacturing

The ones in your fridge that you’re using right now.

But I got them from a reputable lab…

You should sequence your phages — the ones in your fridge that you’re using right now. Word on the street is 12/13 academic labs have mixed phage stocks. Even the most reputable phage labs we’ve all heard of.

But I already sequenced them…

You should sequence your phages — the ones in your fridge that you’re using right now. Phages drift — that’s what they do! You might find that your phage isn’t the same phage as it once was.

But I am careful and I wouldn’t have contaminated them…

You should sequence your phages — the ones in your fridge that you’re using right now. Phages often look like other phages. And they like to hang out in the air and on pipettes and in the nozzle on your vacuum filter that you use to filter sterilize your phages.

But I bought them from a culture collection…

You should sequence your phages — the ones in your fridge that you’re using right now. Culture collections are not perfect either, and even if they sequenced them at one time, sequencing doesn’t always pick up low-level contaminants. One propagation could be enough for that low-level phage to become a majority member of the lysate.

But I’ve kept my phages in the freezer at -80 with glycerol and they literally haven’t been touched since the last time they were opened, and right before we stored them like that, we sequenced them and proved they were clean!

Ok, you’re probably good.

It’s time to stop trusting our phage stocks (or at least: trust, then verify)

It’s time for the phage field to sequence our phages, and get in the habit of resequencing them before we use them in important experiments. Plasmidsaurus exists as a plasmid resequencing service. So does Addgene, in essence. The plasmid people figured this out and they sequence their plasmids before trusting them. Now people buy fresh plasmid stocks from Addgene whenever they have doubts or want to start something new. It saves much more money than it costs.

So why aren’t we doing this with our phages, which are basically plasmids?

But it’s expensive…

It’s less expensive than realizing years of work on a phage was actually years of work on a different phage, or two phages masquerading as one, or that every time you did an experiment it was a slightly different version of this mystery cocktail, and thus you can’t really interpret your data.

And it’s actually not that expensive anymore — it should cost $100 per sample if you submit through Illumina. Plasmidsaurus can sequence them this way - so can many services. These are ubiquitous in most places now.

But it’s hard to get DNA from phages

If you use Illumina you will not need a lot. Plasmidsaurus asks for 20 ul of 10 ng/ul if you use Illumina through them (I used to think they only did long-read, which requires way more DNA, but just learned they also do short read!). Daria Van Tyne’s lab extracts DNA from 250 ul of lysate and it is enough for Illumina. If you use tagmentation you can do it with less than 100 ul (see also here).

But my PI doesn’t want me to spend all my time on this, I have experiments to do

Your experiments will go much faster if you know what you’re actually working with.

But I’m afraid of what I might find...

“I might find out that my lab has been working with phages that aren’t the phages we think they are…”

“My paper came out in Nature and if I find out it wasn’t really those phages in the vials, what will I do?”

“My colleagues/peers/collaborators won’t take my lab seriously anymore.”

Yes, all valid fears. If it helps, I for one will celebrate you for being brave and checking anyway. You can handle whatever comes next — you’ll at least be armed with evidence. It might lead to even more cool discoveries/mysteries solved, or new projects spawned! (A giant chunk of my PhD was based around two very similar-but-not-the-same, spontaneously-varied phages, and why they had different phenotypes).

Let’s just check our phage stocks

Let’s just check our phage stocks. Check the one you’re using actively - just take a sample, extract DNA using any kit, and send it for illumina sequencing.

Use Geneious to assemble the reads and blast to find out what you have. You don’t need to be a bioinformatician. Ask ChatGPT to help you. Map them to the reference genome of the phage you think you have. And maybe map them to a few other phage genomes you work with in your lab.

You might be surprised. It might hurt. Your PI will probably be happy with you in the end, because you might resolve a years-long puzzle about why someone’s project’s results haven’t made sense, or haven’t replicated.

Or you might find your phage stocks are clean - you have 100% match with the phage you think you have. Brilliant! Please store this in glycerol at -80C (follow this paper by the Felix D’Herelle Center team), and make a bunch of aliquots of it in your fridge.

Start a basic spreadsheet: each phage batch gets a number and a location

Start a spreadsheet where you track BATCHES of phages, not NAMES of phages. You now have Phage Batch #1. You know what’s in there. You can confidently share it, do experiments on it, make more of it. You can now do this for all your lab’s phages. If you find a mixture, it’s okay. Just plate it for single plaques on a lawn, and pick each one with a toothpick — propagate from each of those. Now you have more phages than you thought you had! Sequence those ones and bank them once they’re pure.

Don’t get bogged down — just start somewhere

I’ll leave it here for now. Yes, there are many ways to sequence, to assemble reads, to annotate them, to extract DNA, to store phages, to do everything with phages. Just start with one way that gives you a sequence. You can tweak it later. At least you’ll have a window into what phage you have.

How often do I have to do this?

Ideally, every time you propagate your phage, you should check it in some way. However it doesn’t always have to be sequencing!

Easier/cheaper ways to check your phages

Once you have a sequence-confirmed clean stock, there are other ways of flagging cross-contamination more quickly and cheaply. You could establish a small host panel, where each of your phages has a known plaquing profile across them. Like phage A plaques on strain A, B, but not C. While phage B plaques on strain C but not A or B. This is a simple way to flag if you have A or B in your hands. Takes work but if you work with a small set of phages, you could take the time to choose this strain set, demonstrate to yourself that it works, and use it as a sanity check every time you make a new batch of phages. You can also design PCR primers against your phages and run PCR periodically to check that your phages are the same. Or you could do RFLP, where you digest the DNA with some restriction enzymes, and run it on a gel - then compare it to what it looked like the last time (or ideally, when you last sequenced it).

There are many ways to check a phage’s integrity. I am advocating that you do at least something. Or if you don’t, at least know that you might be among the 12/13 phage labs with mystery mixtures or phages masquerading as other phages. And if someone asks you for one of your phages, you might end up having to send it to them without having time to check what it is.

Reputable labs have mixed phages sometimes

It doesn’t make them less reputable — or at least, it really shouldn’t.

As a field, we didn’t always have cheap access to sequencing. We haven’t always known we needed to be this paranoid (and neither did the plasmid people!). We thought (fairly!) that a phage was a phage, and a tube labeled ‘T4’ was T4, and that was that.

And sequencing is still not THAT cheap. It takes time. And phage DNA is hard to get — DNA modifications can make it almost impossible. (FYI: try phi29 polymerase to pre-amplify phage DNA before sequencing if you struggle to get enough DNA from your phage! It will remove the modifications and also amplify the DNA).

But treating your phage stocks with a ‘trust but verify’ attitude — and establishing a freshly sequenced set of phage stocks for each phage you work on (and a process for routinely checking them as time goes on) — is worth the time and money. Think of all the experimental repeats and years-of-grad-student-project-misery that you will save in catching a mystery mixture before its phenotype is declared in a Nature paper, or before it gets mailed to your favourite collaborator.

I genuinely think anyone who works with phages will encounter phage mixtures once they start looking for them. And that’s okay — that’s just phages. But we now have the tools to see them, so we should use them.

I have worked in three phage labs in my life. In two out of three, we found unknown mixtures in our phage vials. In the other one, we never checked…

Acknowledgments

Thanks to Bob Blasdel for presenting on this idea (plus a ton of data!) at Phage Option Colombia last year (and I believe he will give another version of it this year at Evergreen!), and especially telling me about the ‘12/13 labs’ number… ask him about that if you see him!

Thanks to Daria Van Tyne and Jean-Paul Pirnay, who were kind about the unexpected phage mixtures they received from me/my past labs, who kindly sent the data so we could see for ourselves, and shared their favourite phage sequencing pipelines, methods and tips, which I now use (and will write about in the future!).

Cheers to being kind to your fellow phage lab, even if they send you a mixed or mislabeled phage sample!

What are your thoughts?

Does anyone want to add on to this conversation? Do you think I am being too paranoid? Do you want to share an anonymous (or non-anonymous) story of unidentified phage mixtures in your lab?

Or do you have a paper you follow for getting lots of phage DNA that you love? What’s your favourite way to analyze the reads? Do you have a pipeline you could share?

This article was my off-the-cuff rant, but I’d love to compile resources for phage labs looking to get a handle on their phage stocks, and I’d be THRILLED if someone wanted to write a guest post about their perspective on this matter!

Technical Aspects of Phage Research · Standards

As a Canadian living in the US, I've been watching phage therapy developments in Canada with particular interest for the last several years. When Dr. Marisa Azad made headlines (even my grandma sent me the links!) as the first physician to treat a bone infection with phage therapy in Canada, I was excited to one day talk to her.

Thanks to Bradley Cook at Cytophage, I got connected with Marisa, and was lucky enough to have her join as our most recent guest on the Podovirus podcast!

I have always been curious about what the regulatory landscape for phage therapy actually looks and feels like from a Canadian physician's perspective. I was especially curious about the "N-of-1 trial" approach Marisa took for this case (which I blogged about last year ago out of curiosity), versus the compassionate use pathway we see in the US and elsewhere.

In this episode, Dr. Azad walks me through where some of the real challenges lie, the international collaboration she depended on to source the right phage (I was actually surprised to realize she’d indirectly used Phage Directory to send the alert that helped find phages for her patient!), and how she's building reusable infrastructure so future physicians don't have to navigate the same year-long process from scratch.

I hope you enjoy the conversation!

Highlights:

• Regulatory landscape insights: Health Canada (the ‘FDA of Canada’) proved supportive and scientifically informed, while local hospital ethics boards and administrative systems created the most significant delays.
• N-of-1 vs compassionate use: Dr. Azad chose the clinical trial route to create a platform for future patients, rather than treating each case as a one-off emergency.
• Global phage sourcing: She used the Phage Directory alert system to find her patient's phage, which came from Slovenia and was prepared by Cytophage Technologies in Winnipeg — showcasing the international networks essential for personalized treatments.
• Balancing medicine, science, and art: we talked about careers, life, the creativity required to fit it all in, how we find fun in what we’re doing, and the drive to push boundaries so we can help patients.

A few snippets from the episode...

On regulatory realities:

"Health Canada has been excellent. They have really streamlined this process... They're scientists. They've been very helpful. It's really the local regulatory hurdles, which are the biggest challenge."

On choosing the N-of-1 approach:

"I wanted to do a multi-center trial and use this as a platform for that, I wanted to go through the N of 1 or OLIP route or single patient trial route... it would make sense to go more of a clinical trial route if you want to keep going with this."

On global collaboration:

"I literally Googled phage therapy companies and reached out to multiple companies around the world. And it was Cytophage that pulled through... The phage came from Slovenia. It's a small world."

On patient outcomes:

"This patient, thankfully, she has not had any relapses of infection. She has complete wound healing. There's no drainage from the incision sites. She's healthy. She has moved out of her home and she's living independently."

On field collaboration needs:

"Some of my colleagues who've been working in infectious diseases for 40 years, they've commented that the phage world, it's kind of like HIV in the 80s and 90s, it's a wild, wild west. And it's really important that people work together."

What's next for phage therapy in Canada:

Dr. Azad is now leading a multi-center trial platform for phage therapy, expanding beyond orthopedic infections to treat endovascular infections, graft infections, and respiratory infections. Her team is investigating pharmacokinetics, delivery methods like hydrogels, and immune responses to phage therapy.

Meanwhile, momentum here is now building across Canada — just a couple weeks ago, another successful phage + bone infection case, this time in Calgary, made national news.

To me, it is finally becoming clear that Canada is getting on board with phage therapy, and I expect we’ll see continued cases and clinical trials! My grandpa, who’s been telling his doctors for years about phages and demanding why they won’t give them to him, will be thrilled.

— Jess

Listen to the full episode:

🎧 Spotify

🍎 Apple Podcasts

📺 YouTube

Learn more:

• Our interview with Steven Theriault, the Cytophage CEO, on Podovirus: he shares the biotech company perspective on phage therapy regulations in Canada — don’t miss it if you enjoyed this one with Marisa!
• CBC coverage of Dr. Azad's case
• Preprint by Dr. Azad's team on their N-of-1 trial
• Recent Calgary case coverage: Western Canada (my actual home) now has a bone infection treatment on the books with phage too!

Clinical Trials and Case Studies · Regulatory and Ethical Considerations · Phage Therapy Access and Implementation

Welcome to summer, everyone!

This month, Jessica has been hard at work in the phage lab, working with a throng of new summer students. At Groq, I’ve been building tons of open source tools and templates that people can take and run with, modify, and make their own.

This month, I’ve been digging into Google DeepMind’s (of Alphafold fame) latest model, AlphaGenome. I have tons of thoughts on this model that I might just develop into another post. But generally I think it’s a really important stepping stone and research tool, but still basically scratching the surface of what’s possible.

For this month, here’s what we published:

How to navigate regulatory limbo: a Canadian phage therapy CEO’s playbook

by Jessica Sacher

In this Podovirus podcast, Jessica interviews Steven Theriault, CEO of Cytophage Technologies, about his 9-year journey navigating Canada's regulatory maze for phage therapy. Steven shares his experiences building a phage biotech company that has raised $24M and treated patients despite facing regulatory frameworks that don't account for phage biology.

In this episode, they highlight fundamental challenges from how "phages are not drugs," how they don't fit traditional regulatory boxes, how phage variability is seen as "out of spec" (rather than a benefit), to how companies could reach commercial viability through ultra-high titer production that brings treatment costs down to fractions of a cent.

Our Phage Picks for June 2025

by Jessica Sacher

In this month's Phage Picks, Jessica highlights a foundational 2009 paper by Fortier & Moineau on proper phage storage and maintenance protocols, drawing from the Felix d'Hérelle Center's decades of experience managing over 450 reference phages. For Jess, this bread-and-butter technique addresses a common phage library problem where phage collections drift genetically or die in storage due to serial propagation.

Separately, I found an excitingly comprehensive review by Steff Strathdee et. al that surveys the current landscape of AI and machine learning applications in phage therapy, covering everything from infectivity prediction to synthetic phage design. This is a paper I’m already coming back to over and over again!

How to read papers

by Jessica Sacher

In this article, Jessica explores her struggle with efficiently reading research papers. She explains how she’s stumbled across a “3-pass paper reading framework” developed by S. Keshav in 2007. The framework involves three strategic passes through a paper (10-minute filter, 1-hour comprehension check, and deep dive), with decision points at each stage about whether to continue. Jess shared tidbits in a Linkedin which went viral, and she then compiled dozens of responses from researchers sharing their own reading strategies, AI tools, and other frameworks for tackling the endless fire hose that’s scientific literature.

PHANOTATE: Gene finding in phages

Throwback: All this talk about AI nowadays, but we really have to give a special shoutout to to the folks that built the original bioinformatics tools, based on good old algorithms, traditional statistics, and machine learning.

~ Jan & Jessica

Regulatory and Ethical Considerations · Phage Production and Manufacturing · TechBio · Tools

The other day, I spent three hours highlighting a printed-out Nature paper. Digesting it in full — or so I thought. The next day, when my labmate asked me about it, I stumbled through giving her a summary. How is this possible? If spending 3 hours on a paper isn’t enough to understand it, how can I ever hope to get through my reading list, stay up to date on my field, and be successful as a scientist in general?

I went hunting on the internet for better approaches. In this day and age, we should have better ways of digesting papers. Obviously we can now use things like ChatGPT to chat with papers, but I figured there might be more research-centric tools, or even just tips and tricks on HOW other researchers use those tools.

Searching through reddit threads, I stumbled across this 3-pass system by S. Keshav from University of Waterloo. Not some new AI tool, but a 2007 paper from a computer science professor on how he teaches his students how to read papers. I started reading, and really liked it! It was easy to read and understand (I did not even need to spend 3 hours highlighting it), yet had a strict framework about figuring out whether to read a paper, and to what degree, and then tips on how to do it efficiently.

The framework is simple — it involves three passes through a paper, and decisions at each stage whether to move forward to the next or not. I particularly liked this part — how it normalized deciding NOT to read papers sometimes too.

The system made sense, but I had questions: What's the modern version with all our new AI tools? What works for biology, where you can't just "virtually reimplement" experiments, as Keshav mentions (recall: he is a computer scientist, not a biologist like most of us here)? How do we optimize for different research contexts?

I decided to write a post about it on LinkedIn (I did a twitter version too), curious if other researchers were following this method or had thoughts on it.

The response was overwhelming — the LinkedIn post especially went quite viral (as of today there are 490,000 impressions on my post — insane!), with tons of researchers chiming in with their own frameworks, tools, and strategies.

Apparently we are all struggling with this, and yet it seems so many of us have developed our own hacks, workarounds, and frameworks! Most of the posts were people being excited to share their own systems — I loved it!

Keshav's 3-pass system for reading papers

First, here’s the original reference: Keshav, S. (2007). How to read a paper. SIGCOMM Computer Communication Review, 37(3), 83–84. https://dl.acm.org/doi/10.1145/1273445.1273458 (I also found it free online here!): https://web.stanford.edu/class/ee384m/Handouts/HowtoReadPaper.pdf

Here’s the reddit thread where I found it!

And here’s a summary of Keshav’s system:

Pass 1: The 10-Minute Filter (Title, abstract, intro, conclusions only)

Ask yourself: Do I even care about this paper? Most of the time, the answer is no. Move on.

Pass 2: The 1-Hour Comprehension Check

Read with the goal of explaining it to a friend afterward. Look at figures first—sloppy mistakes, missing error bars, or unlabeled axes might signal it's time to find a paper where they actually sweated the details.

Pass 3: The Deep Dive (4-5 hours for beginners, 1 hour for experts)

Only for papers you absolutely need to understand. Go through every assumption and statement. Try to imagine recreating the work yourself. What would you do differently?

Frameworks from the field: what researchers actually do

After I posted my summary of Keshav’s paper last week, along with a few thoughts and takeaways, it started getting hundreds of comments. There have since been almost half a million impressions on the post — for a 2007 paper and some comments on it! Wild times. (FYI my twitter thread on the same thing fared much less well, but still got over 13,000 views, which is a lot for me!)

I thought it was cool that most comments on the LinkedIn post were from people explaining how THEY read papers, and sharing resources and tips — this turned into such a goldmine!

So for this piece, I wanted to pull out and share some of what I found in the comments, including resources, hot takes, trends, and links. But please head to the full post (and add your own two cents!) for the full experience.

Practical tips and frameworks

• Susan Lehman (Phage Biologist):

  "I’m usually reading a paper for something I can’t get from the abstract. […] HOW the authors asked a question is always at least as important as what they took from the answer."

• Rajarshi Ghoshal (ML Science & Engineering):

  "I merge Pass 2 and 3. Read a section, take a break, summarize each in 1–2 lines. If it’s too hard, I beg someone who knows the field to explain."

• Aaron S. (Entrepreneur):

  “Conflicts of interest, funding sources. These tell you in what light to read the paper. Conclusions... methodology... Then cited works.”

• Sean Coffey (Retired R&D Engineer):

  “My advisor also suggested summarizing the relevent details on an index card. He’s had a shoe box full of index cards! I believe the act of writing locks in the details better.”

• Dr. Daniela Palacios López (Thesis Coach):

  “Before going deep into any literature review, you need to have your research question (ONLY ONE), as defined as possible, as well as your methods/methodology, goals and objectives... Ask if the paper addresses the same problem, uses or helps build your methodology, same target group, etc.”(Here’s her Instagram reel on skimming papers)

• Jubilee Prasad Rao, PhD:

  “Read section and subsection headings, the first and last sentences of each paragraph, and understand the images.”

• John Melville, PhD:

  Outlined a 5-step deep dive: assumptions → intro/results → methods → synthesis (do the results really support the conclusions?) → integration with the field (ask where it lands in importance FOR YOU — “Is it a brick in a wall, or is it a keystone arch?”)

Philosophical & meta-comments

I particularly loved the comments that normalized the struggle I was describing in my original post:

• Arazi Pinhas, PhD (Physicist):

  "Modern research is dense and relentless. Spending hours and still feeling lost is normal."

• Robert Evans (Synthetic Biologist):

  "I look for contradictions to what I already know. Then assess: is it me or them who’s wrong?"

• Henry Ward (Software Engineer):

  "If a successful scientist feels stupid reading a paper, maybe the authors need to reconsider their writing style."

AI & tool-based workflows

A bunch of interesting responses came from researchers experimenting with AI tools (of course!), and sometimes combining them with pre-AI frameworks, which I think is a super interesting area.

• Kayleigh Bohémier (Yale Librarian):

  Recommends *Ten Simple Rules for Writing a Literature Review* (I love this series of articles: there are a bunch of Ten Simple Rules papers for everything from doing a postdoc to writing papers to getting grants!).

• Mark Nelson (Stanford Peace Innovation Lab):

  Shared FieldShift, a tool for translating papers across domains to help wrap your head around a paper’s main overarching structure (in terms you’re used to based on your own field, kind of how an analogy/metaphor works), to prime you to understand it better before you dive in. (This I found super cool - you can for example convert an abstract on machine learning into one written as if it were a microbiology paper… or an economics paper. Hard to explain - you need to try!)

• Aditya Gupta (ML Student):

  Uses NotebookLM and Gemini 2.5 Pro to:

  1. Summarize core contributions

  2. Compare papers

  3. Generate study guides and roadmaps

     Also recommends the Google Scholar Chrome plugin.

• Matthias Bohlen (Trainer):

  Uses NotebookLM’s podcast version to hear papers aloud:

  "It’s like a conversation between two voices. Really helpful for early passes."

• Lena Blumenthal (CAS/ACS):

  Uses CAS SciFinder indexing to determine if a paper’s worth reading:

  "The concepts are selected by scientists, not algorithms."

• Samer Attrah:

  Recommends Andrew Ng's Stanford lecture on reading research papers

• Andrew Kightley:

  Shared a toolkit for plain language summaries of clinical trials

Cognitive tricks and note systems

• Thomas Meyer, PhD:

  Uses the Zettelkasten system to organize insights into notes, and extend them with sentence prompts like: “This reminds me of…”, “It’s similar because…”, “It’s different because…”, “It’s important because…”

• Dan R. (Founder, Responsive Humanity):

  Extracts the principle, application, and execution in a bullet structure.

  Example for a paper about a randomized trial on medical student studying effectiveness:

  PRINCIPLE: Spaced repetition beats cramming

  APPLICATION: Med students learn anatomy better

  EXECUTION: Randomized trial, tested 2w and 3mo later

Choosing what NOT to read is just as important

• Mickael Pruvost (PhD):

  “Tip to reduce the "to read" list : If in a paper, there are figures where title, legend, or scale is missing, just blacklist the journal. It means proofreading was missing in the publishing process, and was probably wrote to have a better "author index".”

• Charles Wheelus, PhD:

  "Only read beyond the abstract if it seems worth it. Also note publication date and citation count."

Once a paper gets the green light, some go DEEP

• Brian Greenforest (Engineer):

  “Use paper and pen to map the structure of the original research. Use role modeling and university affiliations. Possible grant funders, try to reproduce the path the scientists have taken to conduct the research. Too much opinionation on your side here, indeed, but it really helps to viscerally sense how they got there. You're re-creating their story. Remember, the paper is a report, you must get to the story first.”

My takeaway

Reading papers doesn't have to be so painful. And none of us are alone in feeling like it’s impossible. There are probably as many ‘frameworks’ for reading papers as there are people reading them…. and I think what matters is having a system that works for you and your brain (and your field).

With so many new AI tools coming out, it’s exciting to me to dig up all the best frameworks that have been written about, and see how we can combine them with the AI tools coming out each day. It is going to be so interesting to watch how our workflows evolve, and whether that leads to researchers feeling more or less overwhelmed. Will it actually give us superpowers? Or will it feed the insecurity that comes from information overload? Will it make our papers more readable, as everyone uses AI tools to ideate and synthesize and write? Will no one read papers anymore if they can just chat with them? How will this influence citation counts — what gets cited, what people are aware of? Will a new form of writing emerge that makes a paper more likely to be something the AI tools like, and are likely to surface to people searching?

So many questions! Ultimately for me in the short term, I especially love the ‘old school’ methods combining with the ‘new school’ tools. I think there’s a ton of great stuff buried in the historical literature, and I want to keep surfacing it and seeing how it can complement our new AI tools.

Also, I mainly just want to stop feeling exhausted and dismayed when I think about how many papers there are to read, and how little time there is… I have hope!

How do you read papers? How many do you actually get through in a week? Feel free to add to the conversation on LinkedIn, where this discussion is still going! I’m trying to respond to everyone! (If Twitter/X is more your vibe, here’s my thread)

Tools · Phage Community and Collaboration · Personal Experiences and Reflections

Hello everyone!

It's Phage Picks time! For my pick this week I want to go back to another foundational method (a bread and butter technique paper that every phage lab should master).

My pick might not be the flashiest, but could be one of the most important for promoting reproducible phage research. It's about properly storing your phages so they don't evolve away from you over time, written by some of the curators of the Felix d'Hérelle Center (who’ve arguably been keeping more diverse phages alive for longer than any of us!)

After that, Jan’s pick this week is about — try to guess — phage and AI! But this is a new perspective that gathers what all has been done with phages and AI to date, and even Jan said there’s a ton in there he hadn’t heard of. His reading list is bursting! If you want an overarching survey of phage+AI, this is what you need to read.

Hope you’re all having a great start to summer! We are working away at our respective Real Jobs, loving the fact that summer is basically year-round here in Palo Alto, and I’m finally reunited with my horse after pandemic years + Australia years kept us apart. The lab at Stanford is starting to swell up with summer students again, and the vibe is great!

Cheers to good weather and continued grant funding for all,

• Jess

Phage Production and Maintenance of Stocks, Including Expected Stock Lifetimes

What is it about?

This is a ‘how to store your phage’ paper from 2009 that I keep coming back to; I even printed it out today (long overdue). Why? Because in each lab I join, I feel and witness the same struggles with phage collections that have drifted genetically or simply died in storage. Or worse, it inevitably happens where someone’s not sure which version of a phage they used for a given paper, because they didn’t archive master stocks of each. Instead, maybe everyone’s been serially propagating from the stocks over years. Do multiple lineages of each phage exist in the lab now? Probably.

But this shouldn’t be a new problem; there should be some tried and true methods to deal with this by now, right?

Here’s your handbook! Simple, straightforward advice for at least the first step of this issue: archiving your phages so they last, so you DON’T have to serially propagate them to keep them going through the years.

Why I’m excited about it?

First — because of who wrote it. The authors of this paper manage the Felix d'Hérelle Reference Center for Bacterial Viruses at Université Laval, so they have access to over 450 reference phages and 20+ years of storage data. So this is not theoretical advice, or based on the experience of one or two model phages, it's a set of battle-tested protocols and important concepts from people who've kept hundreds of different phages alive for decades.

The authors describe protocols and tips for phage storage across multiple temperatures/conditions, including refrigeration at 4°C, deep freezing at -70°C and -196°C, and lyophilization (freeze-drying).

Some great tips and observations emerge, like how they found that vacuum quality matters most for long term lyophilized phages, or how different phage families respond dramatically differently to storage methods. For example they saw that large phages (Myoviridae like T2, T6) suffered major losses during freeze-drying (up to 3-log reduction), while smaller phages (Siphoviridae like T1, Microviridae like φX174) were much more resistant (less than 1-log loss).

They also lay out some nice quality control frameworks to use to monitor your phages (electron microscopy, host range testing, sequencing, but also a simple ‘pick 6 plaques and do a restriction digest on each’ method - if they’re all the same, move forward; if they’re not, you have a mixture). The latter could come in very handy if sequencing isn’t readily available; a classic technique I feel we don’t often think to use.

I also like their emphasis on solid metadata collection and database management, something we’ve had top of mind at Phage Directory for a long time (and which I’m now helping implement in my current lab).

Overall, I think with some simple but strict practices (no more serial propagation! no more!), we can build stable phage libraries that maintain their properties for decades, and have all of our great phage research actually compound on itself (instead of having to basically start over every time a phage gets passed to a new person, or someone graduates).

I hope to write a full guide on this in the future! For now, just check out this paper.

Paper: Fortier, L. C., & Moineau, S. (2009). Phage production and maintenance of stocks, including expected stock lifetimes. In Methods in molecular biology (Vol. 501, pp. 203-219). DOI: 10.1007/978-1-60327-164-6_19

Link: https://pubmed.ncbi.nlm.nih.gov/19066823/

— Jess

Optimizing phage therapy with artificial intelligence: a perspective

What is it about?

In this perspective article, Steffanie Strathdee and team wrote a really comprehensive review of the phage therapy field in terms of applying AI and machine learning. They did a really good job surfacing all the various advances across modalities of phage prediction, e.g. can we use ML to predict phage infectivity? Can we identify novel phage genes with therapeutic functions? And what about designing synthetic phages?

The paper goes fairly deep into a vast number of phage/ML projects that have been done in the last couple of years. If you’re looking to get caught up on phages and ML, this is the paper to read. (There are a ton of papers I’d never even heard of before!)

Why I’m excited about it?

The article goes quite deep and accurately (in my opinion) identifies the wider problem that we’re able to predict phage infectivity for various strains, but lack a more generic large infectivity model across species. Current models use clever insights about specific genus or strains, but when it comes to more complex pathogens like S. aureus or P. aeruginosa, it gets much harder — with larger diversity in receptor types etc. It also correctly identifies a lot of potentially narrow overfitting problems in models since they’re trained on relatively small datasets.

The article boils down to — most phage/ML projects are biologically informed, meaning they are designed around our base understanding of biology.

The paper doesn’t mention this, but it seems like the field seems to still be lacking efforts in building unsupervised and/or self-supervised models that try to tackle everything, from phage/bacterial genomes, to annotation metadata and phenotypic rock-paper-scissors kinds of interactions; basically a foundation model for phage/host and receptor-binding protein prediction (or maybe even synthetic phage genome prediction??).

The answer to why haven’t we built a Phage Foundation Model (a Phoundation model??) is probably easy — it requires a ton of data (which is hard and expensive) and it needs a ton of wet-lab validation (which is also hard and expensive). But it’s not difficult. It’s fairly straightforward. It just requires lots of patience, hard work, and loads of funding.

— Jan

Paper: Doud, M. B., Robertson, J. M., & Strathdee, S. A. (2025). Optimizing phage therapy with artificial intelligence: a perspective. Frontiers in Cellular and Infection Microbiology, 15.

Link: https://www.frontiersin.org/journals/cellular-and-infection-microbiology/articles/10.3389/fcimb.2025.1611857/full

Phage Production and Manufacturing · TechBio · Technical Aspects of Phage Research

"Phages are not drugs. Every time they say, 'Did you go through regulatory?' I say, 'I can do regulatory, but I'm not a drug.' There's 145 components of the regulatory requirements that I don't fit in."

When your health innovation doesn't fit existing regulatory boxes, how do you build a business? Steven Theriault, CEO of Cytophage, has spent 9 years learning to navigate Canada's regulatory maze for phage therapy. From being told "we don't know" by government officials to raising $24M and treating patients, Steven shares his hard-won playbook for building in uncharted regulatory territory.

In this episode of the Podovirus Podcast, we talk to Steven Theriault about what he's tried, accomplished and learned in the last ~decade building a phage biotech company in Canada:

Highlights from our conversation:

• The "not a drug" dilemma: Traditional GMP guidelines don't account for the inherent variability of biologics – if Steven's phage grows to 10^12 instead of 10^10, that may be better for treatment, but regulatory frameworks would see it as "out of spec"
• Regulatory geography matters: Cytophage has more traction with U.S. and international regulators than in Canada, despite being a Canadian company, because other countries are adapting their frameworks for phages
• The emergence of new regulatory hurdles: Mid-treatment, a new government organization emerged requiring four months to review phage preparations – potentially fatal delays for dying patients
• Commercial viability through reduced cost per dose: By achieving 10^15 phage titers with treatment doses of only 10^3, Steven's team can achieve a treatment dose [for chickens] for something like 0.00006 cents — well below the pennies-per-dose that antibiotics go for

A few snippets from the episode...

On leaving government to pursue phage innovation:

"I realized that the government's not necessarily the best place for out of the box thinking. So for myself, I stepped out of the government, again, very secure job, worked there for about 15 years, had a great career, stepped out because I really saw a need as well as something that we could do with phages."

On the reality of treating patients in Canada:

"For human, we've always looked at human sides. The biggest thing for us though is we can only do about four patients a year. And the reason we only do four patients a year is because of cost restrictions. If this was a funded program, I could do a hundred patients a year."

On the disconnect between science and regulation:

"Listen, we all know toxicity is not an issue. When I put in my first application in the Canadian government, of which I have four currently, applications have been sitting there for 18 months, they asked us to come back and show toxicity. So I went ahead and I paid for a study, and this is like the most boring study you'd ever read in your life. Nothing happened."

On the funding momentum shift:

"I haven't been this excited about phage research and the success of phages since I started this company. It's just this new year that sort of really showed that people are starting to flip the switch and really understand."

What's next for regulatory reform:

Steven is working on developing guidance documents for phage research in Canada and expects significant policy changes with the new government. His vision includes coordinated phage treatment protocols across Canadian research centers and potentially a dedicated phage research facility to streamline patient access.

Listen to the full episode:

• Spotify
• YouTube
• Apple Podcasts

Learn more:

• Cytophage website: Learn about their agricultural and therapeutic phage applications
• Steven's TEDx talk on the future of phage therapy
• CBC story on Thea, Cytophage (and Canada’s) first PJI patient treated with phages

This episode is part of our ongoing series exploring the practical challenges of bringing phage therapy to market.

Regulatory and Ethical Considerations · Intellectual Property and Commercialization · Phage Production and Manufacturing

Hi everyone,

Spring is upon us! Jess has been focused on her research — a slight deviation from phage therapy into phage therapy — using phages to deliver large genetic payloads as gene therapy. Is that still considered phage therapy? The gene therapy conference was a success, and she’s been going down a few rabbit holes in this new, tangential realm.

Coincidentally, a new article was published in Nature, around the “pivot penalty” in research. Take it with a grain of salt, but from the authors’ analysis of millions of research papers, they conclude that “the further a scientist moves away from the topic of their previous work, the fewer citations their new work will receive.” The paper seems all kinds of flawed, but there’s probably some truth to it. Pivoters, beware. Read more on Nature: https://www.nature.com/articles/s41586-025-09048-1

As for myself, my work at Groq is getting on-track and I’m getting in the groove. I’m back to prototyping and building AI-based research tools. Most of them we still use internally (e.g. I share with Jess to use) as they’re still quite janky, and break a lot. When they work, they really work though. We use some of these tools for Capsid research, and to write issues like Capsid & Tail. My personal goal is to make these reliable enough for Jess to use on a daily basis in her research — at that point I think it would be ready for release!

For this month, here’s what we published:

Finally, phase 2 phage therapy data! Inside BiomX’s successful trial with CEO Jonathan Solomon

by Jessica Sacher, Jan Zheng

In this Podovirus podcast recap, Jessica Sacher interviews Jonathan Solomon, CEO of BiomX, about their successful phase 2 clinical trial for diabetic foot osteomyelitis using phage therapy. The trial showed not just microbial effects but actual clinical improvement with statistical significance, with approximately 40% improvement in the phage-treated group and impressive tissue regrowth in patients where bone was exposed. Their approach used a single optimized phage for each patient rather than a cocktail, and demonstrated that targeting only Staph aureus in polymicrobial infections was effective.

Why can’t more patients access phage therapy?

by Jessica Sacher

In this Podovirus podcast recap, Jessica interviews phage therapy patient advocate Chris Shaffer about his journey to access phage therapy, and his ongoing advocacy work.

After successfully treating his persistent E. coli infection at the Eliava Institute in Georgia, Chris now helps other patients navigate the challenges of awareness, trust, and access to phage treatments. He draws parallels between the current situation and the early HIV/AIDS crisis, suggesting that patient advocacy groups similar to ACT UP might be needed to increase public demand for phage therapy.

Finding papers at scale

by Jan Zheng

In this post, I outline a few strategies and tools for finding relevant research papers. There’s too many papers getting published (and coincidentally, also too many tools “that use AI” to read those papers) right now. I’ve basically tried them all, and I’ve summarized them here. Some tools are better for finding the right paper to read; others are better for discovering new papers — I lay out all my favorite ones here. Most of them have a free tier, so there’s no excuse not to explore!

Why aren’t we using primerless PCR for microbial discovery?

by Aaryan Harshith

In this article, Aaryan, our regular guest writer, explores how pipolins (genetic elements that can replicate DNA without primers) can be used for isolation. Current isolation methods are too constrained and selective, so we’re unable to detect most species in our samples. Aaryan then outlines a protocol for implementing pipolin-based PCR for a microbial discovery pipeline. If you’re interested in pipolin-based PCR (pPCR?) you should reach out!

How do we commercialize phage therapy? A full recap of Phage Futures 2019

Throwback: For better or worse, it’s been ~5 years and the field is still grappling with this question. With several positive results from BiomX/APT and Armata clinical trials, and some large shifts in the biopharma/biotech realm, this question is probably more pressing than ever!

~ Jan & Jessica

Clinical Trials and Case Studies · Phage Therapy Access and Implementation · Tools · Microbiome

Disclaimer: Phage biology is a rapidly evolving field. All the data and literature I present in this piece are current as of May 14th, 2025. If you’re reading from the far future, I always recommend checking the latest primary research.

Discovering microbes isn’t some arcane pastime—it’s the heart of microbiology.

Flip through our scientific timeline, and you’ll notice microbes everywhere.

From Iceland to India, yeast have helped humans ferment for thousands of years. Bacteriophages enabled the discovery of CRISPR. Even molds led to the discovery of penicillin, enabling a generation of antibiotics that extended the human lifespan by nearly two decades (Hutchings et al., 2019)

The pattern is strikingly clear. Some of humanity’s most impressive biological triumphs— from vaccines to restriction enzymes—were either enabled by the discovery of a new microbe or inspired by its unique biology.

Dr. Thomas Brock, whose discovery of the Thermus aquaticus in Yellowstone National Park enabled the development of PCR. Image courtesy of NYT.

And yet, as one of the most exciting frontiers in science, microbiology remains shockingly unexplored. In one review, Rappuoli et al. (2023) estimated that our planet harbors at least a trillion unique species of microbes. But scroll through NCBI’s dataset of microbes and you’ll find ~100,000 species — less than .00001% of what exists.

If our scientific empire is built on such a tiny sliver of the microbial world, what possibilities are we missing out on? And more importantly—what’s stopping us from finding more microbes?

As it turns out, there’s a reason for this. It’s less that we don’t want to and more that we’re bottlenecked by our tools.

Today’s methods of isolating microbes are inherently selective. When we hunt for viruses, we select against a host. When we culture bacteria or fungi, we grow them in media that nurture some cells and starve others. If we want to enable the microbial renaissance, we need a truly agnostic test — one that can identify microbes indiscriminately and at scale.

Meanwhile, at the Pasteur Institute, Mart Krupovic’s lab has discovered a novel element in bacterial genomes capable of replicating DNA without primers. While these modules, termed pipolins, have shaken a basic axiom of DNA replication, they’ve largely been sidelined as a scientific curiosity. (After all, why would anyone want a non-specific DNA polymerase?)

And yet, it seems like we’ve overlooked the most interesting application of pipolins — microbial discovery. The same machinery that lets pipolins replicate DNA could enable the world’s first primerless PCR. Effectively, this would let us amplify trace genomes from bacteria, viruses, and parasites — regardless of how little we know of their sequence or structure. It would offer the first complete snapshot of our microbial world.

Our best science has always emerged when a compelling idea finds its proper use. Pipolins are up next.

Hunting for Microbes

There are already many ways to catch microbes in the wild. Most methods either resort to growing microbes from the environment, or directly sequencing their genomes.

However, as I demonstrate in my earlier piece, both systems suffer from the microbial analog of selection bias. Even the simple act of streaking soil samples on a plate filters bacteria—in this case, based on their ability to metabolize nutrients in the media¹. Today’s methods of isolating microbes are inherently selective, and therefore exclusive.

On the other hand, techniques like metagenomic sequencing let us probe entire microbial communities, but only when there’s enough genome from each resident. With the current detection thresholds of metagenomic sequencing, we’re likely missing out on trace microbes that make up most of our microbial diversity.

The basic isolation protocol for human gut (enteric) microbes. Image courtesy of Wan et al., 2023.

We can amplify microbes with PCR, but only when we have specific primers for their DNA. But this quickly leads to a chicken-and-egg dilemma. To amplify a microbe, we need its sequence. But in order to sequence, the microbe already needs to be amplified.

While techniques like these are handy for isolating microbes with known properties, they’re blind to everything else. But what if we weren’t looking for any microbe in particular? What if we wanted to cast the widest net possible and narrow down later.

That would call for a fundamentally different net. And this is what pipolins will let us build.

The Lost World of Pipolins

Long ago, back when biology was buzzing with the discovery of DNA polymerase, biochemist Arthur Kornberg isolated a curious protein called DnaG.

With further experiments, Kornberg showed that DnaG didn’t string together novel DNA, but instead laid down short scaffolds of RNA where DNA polymerase begin replication (Rowen and Kornberg, 1978).

Since then, our knowledge of primers has flourished. In animal cells, primers are built with RNA. In certain viruses, proteins like VpG shoulder this role, latching onto the genome and serving as a foothold for replication to begin.

While the form and process differ, priming itself was always seen as universal. At least, until recently.

In 2017, Mart Krupovic and his group at the Pasteur Institute published a paper that quietly fractured this assumption. While comparing polymerases across microbial clades, the authors identified a novel genetic element. After isolating the resulting enzyme, they found it had the strange ability to replicate unprimed DNA, and decided to call it primer-independent polymerase B (pipolB).

Further exploration revealed this wasn’t a stroke of luck—similar genetic modules were found across the bacterial kingdom, in certain phages, and stowed away in symbiotic organelles like the mitochondria.

The hallmark of pipolins is their capacity to polymerize DNA de novo, without RNA or protein primers. Image adapted from Redrejo-Rodríguez et al. 2017.

To understand this better, I’ll let the authors explain directly:

Collectively, these results indicate that piPolB from E. coli 3-373-03_S1_C2 pipolin is able to initiate and perform DNA replication of circular and linear templates in the absence of pre-existing primers or additional protein factors.

Considering how radical this finding is, it’s a surprise it hasn’t echoed further throughout the scientific community.

When you follow this logic, it opens up an incredible possibility for microbial discovery. With the unique capacity of pipolins to join nucleotides from scratch, we could design a non-specific PCR — one that could amplify all the microbial genomes in a sample. By enabling replication without exact primers, this approach could capture the presence of microbes that otherwise go undetected.

Of course, this makes some major assumptions. Let’s think through them in single file.

Q: But don’t microbiologists already have tools to amplify all the genomes in sample? Isn’t this a relatively common approach?

It is true that certain fields in microbiology (specifically gut microbiology) have methods to replicate genomes from entire microbial communities.

The most common technique, known as whole genome amplification (WGA), relies on massive libraries of random primers in hopes of initiating DNA replication everywhere. It’s the genetic analog of a brute force search.

Mathematically, this doesn’t make much sense. Depending on the genome, primers can range from 18-30 nt. Even assuming the smallest size, a truly comprehensive library would need >60 billion unique primers². Our largest libraries today range up to 24-mers, meaning we still miss out on microbes since we don’t cover the full space of primers.

Even if it did, combinatoric chemistry is expensive. While WGA excels at one-off tests, it quickly becomes impractical as we scale to thousands of samples — let alone the millions we’ll need to map the planetary microbiome.

This is why pipolins aren’t just a convenience. By doing away with the priciest piece of genome amplification (namely, our giant libraries of primers), pipolins could let us replicate DNA as long as we feed our system with dirt-cheap dNTPs and ATPs. It’s a radically different technique since it can scale to unseen volumes of microbes.

Q: But aren’t most polymerases sequence-specific? Do we have any idea if pipolBs can replicate DNA aside from the sequences they normally act on?

Fortunately, they do.

In the original paper, Krupovic’s group showed that pipolB could initiate replication across a giant space of sequences — from native DNA in E. coli to fully-synthetic 33-mer strands of deoxythymidine.

Electrophoresis shows that pipolins can replicate both primed and non-primed DNA from phage M13. D368A is a non-functional mutant of the wild-type pipolin.

Image courtesy of Redrejo-Rodriguez et al. 2017

To illustrate, here’s another quote from the paper:

Furthermore, replication of homopolymeric DNA substrates suggests that, contrary to canonical DNA primases (Frick and Richardson, 2001; García-Gómez et al., 2013), piPolB DNA-priming capacity does not rely on a specific template sequence.

For now, it seems that if the target molecule is dsDNA and it’s structurally accessible, pipolB can replicate it.

Though it’s tough to prove that pipolB acts on all DNA, it’s reasonable assume that it operates on an incredibly broad sequence space.

While non-selective polymerase would be a bug in any other context, it’s exactly the feature we need for microbial discovery.

Q: Even if we could use pipolins to replicate mixed populations of DNA, could we ever assemble reads from so many unique genomes?

As enzymes, we expect DNA polymerases to function as long as they operate in the proper environment (i.e. pH, temperature, cofactors), and are constantly replenished with nucleotides.

If anything, the limiting factor is our ability to assemble DNA.

Currently, our best algorithms to assemble reads from metagenomic sequencing (i.e. SCAPP, Meta Flye, etc.) still need coverages of at least 20-fold to function properly (Ni et al. 2023).

Consider a lone bacterium. If, as a crude estimate, we assume a genome of 5Mbp, replicating this 20-fold is already a massive feat³. Now consider a complex microbial community with billions of bacteria. At least intuitively, this seems even more challenging.

Metagenomic sequencing is a challenging feat with many approaches. Image courtesy of PacBio.

Physically, there must be some upper bound to how many microbes we can resolve in a community. In any case, if there’s a bottleneck in this process, it likely won’t be the polymerase, but the number of genomes we can fit in solution and assemble successfully⁴.

From a purely biological perspective, Krupovic’s paper rebuts a core assumption about DNA replication. And yet, a quick search on Google Scholar shows that this paper received only 34 citations. At least for the time being, the community seems to view pipolins as a scientific curiosity.

But rightly so.

Historically, DNA primases have glimmered because of their specificity. When we first unraveled their function, our most earliest and most impactful solutions used it to selectively amplify DNA. Think about approaches like PCR, which let us identify bacteria, viruses based on particular sequences.

It’s quite natural that the idea of primerless replication hasn’t caught on. What use could anyone possibly have for non-specific DNA replication?

In practically every use case, it makes sense to seek out precise sequences of DNA. But microbial discovery could be the most important exception.

In this singular, yet incredibly meaningful edge case, this quirk of pipolins isn’t a bug, but a feature. When we scour the planet for genetic footprints of unknown microbes, we want to amplify everything indiscriminately.

Perhaps no one reading this paper fancies hunting microbes. Pipolins might be exactly what this field was waiting for.

Revamping the Microbial Discovery Pipeline

As I mentioned earlier, microbes today are largely discovered through brute force. We begin by collecting samples from the environment, isolate microbes using mostly arbitrary selection filters, then sequence our results. While this approach is routinized and reliably churns out new species, it is fundamentally wrong. What we isolate simply isn’t representative of the world.

Because of their inherent design, we aren’t seeing all microbes in the planet — merely the small fraction that are responsive to our tests.

If we want to resolve the microbial world (and maintain our sanity) we can’t use a piecemeal approach. We need an agnostic system that can identify all microbes in a sample.

It wouldn’t take much to integrate pipolins into microbial discovery. Here’s one protocol I designed from my work isolating viruses.

Consider a soil sample. After suspending the sample in an arbitrary buffer (i.e. SM, PBS, etc.) and filtering through a 0.2 micron membrane, we’ll have a solution packed with viral particles and small molecules. Treating this with a lysis buffer will give us a batch of exposed DNA.

Basic workflow for microbial discovery with pipolins. Following treatment with pipolins, the process mirrors standard PCR.

Alternatively, if we wanted to resolve RNA (as is often the case with animal viruses), we would simply use RT to turn the entire sample into DNA first. Then, much like PCR, we would heat the sample to de-anneal DNA and treat it with pipolB to begin the replication process.

Since pipolins are distinct from normal polymerases, they have unique enzymatic needs. While current literature suggests that pipolB functions optimally with manganese (Mn²⁺), these parameters can always be optimized with concentration testing.

From here, we would follow the standard PCR heating cycle to amplify all the genomes in our sample.

Then, after sequencing with standard techniques (i.e. Illumina, PacBio), we could then resolve our assembled reads.

As you can see, the bulk of the process remains the same. The end result is still snapshot of a microbial community— the only difference is that we now have a panoramic view.

And Now, Your Turn

As a virus hunter, I believe that microbial exploration is one most important scientific pursuits.

If you’re reading this article, I assume you care about this, too.

Microbiology has already changed our world. Now we’re nearing a precipice where microbiology becomes our world and makes it larger.

The community of microbiologists is still small, and that gives us the chance to build the tools that will define this field for generations.

Over the next few weeks, I’ll be in touch with the Krupovic lab to see how our ideas can intersect, and potentially running experiments with pipolins on phage DNA.

This article is an open invitation. If you want to help design and test this approach, write to me and we can find ways to collaborate.

Let’s make this real.

Footnotes

1 - And no, this distinction isn’t just wordplay. Some of the most clinically relevant bacteria today (including spirochetes responsible for syphilis) weren’t studied for decades because they couldn’t be grown on traditional growth media. This is the drawback to selective tests — they assume that what we see is all there is.

2 - This would be a conservative estimate since we ran four to the eighteenth power, which models the combinatorics of our shortest primers.

3 - Here, I use the term “bacterial genome” to illustrate the downstream effects of continuing DNA replication. For viruses, fungi, protists, or other microbes, we would obviously expect a similar effect depending on the genome size of each microbe.

4 - If there is an upper limit to the number of unique genomes capable of assembly, it doesn’t seem like related fields (i.e. gut microbiome mapping) have documented reaching this despite using whole-genome amplification.

Microbiome · Technical Aspects of Phage Research

I’ve never been particularly fast at reading papers. Usually, just skimming a research paper takes me about an hour. Reading a paper — by trying to understand it “completely” — can take me at least 4-5 hours. And it leaves me exhausted at the end.

Most papers I’ve read cite at least 10+ papers. ~5M papers are published per year. If someone were to “read everything that’s ever written” about a topic, they’d never do anything else. And the worst part is, spending all this time reading a paper, then finding it’s irrelevant.

There are tons of great papers out there, but to cut to the chase, we need to focus only on those that are relevant and worth our time. While Pubmed and Scholar are still invaluable, here are a few new search tools that can help you figure out which rabbit holes are worth going down, and which you can skip.

I’m at a GSK talk while writing this, and to my surprise, they’re explaining how they use AI to speed up their paper discovery and hypothesis loop. Ideally, their flow is: Exploration > Literature synthesis > Hypothesis > Update knowledge > More Exploration. With these new tools, they can scale up this process to hundreds of thousands of iterations. They’ll then review the final stack for the most interesting ideas.

While the following tools aren’t at GSK’s level, they can still help your lab approximate this flow. Below are some of the tools that Jess and I like to use for research and reading tasks. (We’re not affiliated with or get paid by any of them — we’re just fans!)

Elicit.com is designed for researchers, and generates a really nice table of all papers relevant to the search term. Jess and I have been Elicit fans for a very long time!

Science-specific search tools

We’ve all used Google, Google Scholar, Pubmed, and Mendeley for regular research work, but here is a selection of new search systems specifically geared for scientific research. Think of these as chimeras of Google Search, Mendeley and other science platforms. These platforms are meant for performing an array of tasks like lit reviews, data extraction, and full research tasks.

Elicit - http://elicit.com - I’ve spent quite a bit of time with Elicit, and this is my favorite tool for digging deep into one specific topic (e.g. phages and optogenetics). It searches a graph representation of related papers, and can summarize an entire thread or domain fairly quickly, surfacing exactly the papers you need to know.

SciSpace - free + paid tiers - https://scispace.com - Similar to Elicit, SciSpace is a system for searching for papers along queries and topics. It also doubles as a powerful reference manager, and has AI features like summarization. I commonly will check results between Elicit and SciSpace.

Consensus.app - free + paid tiers - https://consensus.app - Consensus is a clever take on searching science. Designed more for laypeople, it takes statements like “does phage therapy work” shows search results both for and against the question, and figures out where the consensus lies. This one is excellent for writing more accessible articles for science communication and similar tasks.

FutureHouse Platform - free - http://platform.futurehouse.org - FutureHouse is a nonprofit for building science automation tools. Their new Platform is a framework for comprehensive lit search, critical analyses, and hypothesis validation. Eventually, they’re aiming to build a full “automated scientist” where they can pair up both data and lab in a closed loop. Learn more about their work here.

The Future House Platform has various nuanced search systems for researchers. https://platform.futurehouse.org

General search tools

There’s also quite a few AI-native search platforms that have been released in the last few months. These systems summarize, contextualize, and extract data from hundreds of search results. Below is a short list of the search tools that I’ve tried.

Kagi Search - free + paid tiers - https://kagi.com - Kagi is not an AI search tool. It’s an old-school query-based search like Google, but unlike Google it doesn’t have ads. Since they’re not incentivized to show sponsored results, the results are generally much better than what Google shows. They do have their own version of Deep Research though, but it’s still in beta.

Google Gemini Deep Research - free + paid tiers - gemini.google.com - Invented the concept of Deep Research. Runs dozens to hundreds of Google searches, finds relevant papers and links, and cites the sources as it answers the user’s questions. This is very good for terms that are otherwise hard to find on Google, but can take several minutes. You can also look up papers related to a given paper, or do a quick check if anyone has ever thought of a specific hypothesis. This alone will save you many hours a week.

OpenAI Deep Research - free + paid tiers - openai.com - Very similar to Google’s Deep Research, and works somewhat similarly. I think they use Bing Search instead, but otherwise the results are similar.

Perplexity - free + paid tiers - https://perplexity.ai -  This is a completely new search service, and can sometimes have better integrations with Pubmed and other paper databases. They have standard search and deep research, as well as a feature that creates “wikipedia”-like pages of results. I find their search is faster, and seeing how it searches for various papers, “thinks” about the results, then continues to search, is really neat. I normally use Perplexity as an alternative to Google.

Anthropic Claude Search - paid - https://claude.ai Claude is an underdog in the AI field, and excels at writing prose and code. We use Claude often for paper summarization. Claude Search is a new feature, and while it doesn’t give as detailed reports as Gemini or OpenAI, it does create better writing outputs.

NotebookLM offers a way to chat with papers, plus you can use it to create engaging podcasts! notebooklm.google.com

Audio research tools

Listening to a paper quickly helps me get the gist of a paper, and helps me find all the parts I should pay attention to, before sitting down and diving deep into it. I highly recommend listening to any paper you’re seriously considering fully understanding.

NotebookLM - free + paid tiers - https://notebooklm.google.com - Creates a podcast from (and lets you chat with) any paper, PDF, or report. The ~10 minute podcast is great for quickly grasping papers from areas I’m not too familiar with (e.g. machine learning)

Listening - $13/mo - https://listening.com - Creates a full narrated version of any paper. Jessica uses (and pays for) this one. The audio isn’t perfect, it doesn’t handle images or tables, and sometimes it breaks in funny ways. But it’s still great for long commutes.

Groq’s experimental Deep Research, Fast, which creates a research report in ~10-15 seconds https://deep-research-fast.vercel.app

Programmable Search Tools

For researchers and bioinformaticians who write code, there are a few other notable tools that can mostly only be accessed programmatically.

Exa - Free + paid tiers - https://exa.ai - Can run multiple, parallel searches, and can be useful for those building a research summarization tool. The Websets feature does research in the background and can dig up thousands of relevant papers and other sources on a topic.

Jina Reader - Free + paid tiers - https://jina.ai/reader - Is very good at reading and taking screenshots of websites, but can struggle with research sites. While it can’t access paid journals like Nature, it can access most open source journals.

Groq Compound - Free + paid tiers - https://console.groq.com/docs/agentic-tooling/compound-beta - lets you build advanced web search and code execution through a single API call. This makes it easy to build your own custom versions of Deep Research and AI Scientists, customized to your own needs and your own data. (Side note: this is where I work now!)

Bonus tool: a super fast version of Deep Research, powered by Groq: https://deep-research-fast.vercel.app

One last note

That’s a lot of tools for search!

Whichever option is better, is completely subjective! Since they all have a free tier, there’s no excuse not to try a new one every week, and then compare for yourself. Given how large the potential benefit is of trying one of these tools, it’s really worth the 30 minutes to try them.

These tools are free, they can shave off countless hours of research, and they are free to start, so there isn’t really any excuse to not try them.

Happy reading!

~ Jan

Tools · TechBio

"This feels just like the early days of the HIV crisis. People are dying, you're not hearing about it. We need a group like ACT UP to bring this to the public. We're not going to get phage therapy until people start demanding we have it."

This week on the Podovirus Podcast, we talk to phage therapy patient advocate Chris Shaffer about how he fought for access to phage therapy to save his own life, and how it's led him to advocate on behalf of other patients being told there's "nothing left" for their infections.

He draws parallels between phage advocacy and 1980s AIDS activism, while sharing his successful treatment journey to Tbilisi, Georgia for phage therapy. He shares his vision for a patient-centered future where anyone, not just the well-connected or the lucky, can access phage therapy if it could help them.

A taste of what we discussed:

• First: Chris’ own clinical success: After two years of consistently high bacterial counts (>100,000 CFU) while on antibiotics, Chris saw his E. coli with ESBL infection completely eliminated after just three 20-day rounds of phage therapy at the Eliava in Tbilisi —with no side effects and sustained clearance 20 months later.
• Information barriers for patients: As Chris sees it, phage therapy still faces a three-part challenge: awareness (patients don't hear about it), trust (patients aren't sure it's legitimate), and access (even informed patients struggle to find treatment)
• Medical education gap: An infectious disease intern Chris met from a prestigious US institution had never heard of phage therapy despite just completing medical training—underscoring the systemic nature of awareness barriers.
• Patient-to-patient networks: Before committing to treatment in Georgia, Chris needed to connect with another phage therapy patient (Pranav Johri) who had been through the process—highlighting how patients might solve the trust barrier for each other.
• Potential fundraising model: We discuss how small-scale patient fundraising networks could provide the modest funding needed for academic labs to prepare phage treatments for compassionate use cases, similar to Diane Shader-Smith's fundraising and advocacy work on behalf of her daughter Mallory.

A few snippets from the episode…

On the communication challenge:

"Obstacle number one, you will not hear about phage therapy. It doesn't matter what doctor you're seeing in the United States, you're not going to hear the word phage therapy... If you are able to get a compassionate use waiver to try phage therapy in the United States, the next obstacle is finding the phage. I didn't know anybody. Who am I going to turn to to find a phage that's going to work for my infection?"

On his own infection recovery:

"I woke up in the morning on the third round, the 12th day, and I walked into the bathroom to do my morning routine... I was looking at my weathered face and I said to myself, 'It's gone. It's literally gone.' I could tell. I didn't need a test... I sat down on the edge of the tub and I started to cry. I literally did because I knew it was gone. And, sure enough, I finished that last round of phage. Two weeks later, I had a test. No sign whatsoever of E. coli, and it has remained that way ever since."

On his unique advocacy position:

"I sit in a very unique place. I'm 67 years old, I'm retired, I have a pension. I don't need any money. I don't have to look for a job. I'm not doing this because I think I'm going to get some kind of job out of it... I can say exactly what I want to say. And not everybody has that freedom... I can't wait until the day where they say Shaffer, nobody needs you anymore because phage is out there, everybody knows about it and everybody's using it."

What's next for phage advocacy?

Chris envisions several pathways forward: establishing patient-funded compassionate use centers, better educating physicians starting in medical school, and creating clearer communication about the FDA's stance on phage safety. He's exploring the possibility of a Palm Springs treatment center that could serve as a model for patient access, and suggests developing small-scale fundraising networks to support research labs that make phage preparations for compassionate use. His approach balances respecting the scientific process while arguing that existing regulatory pathways could be better utilized for greater access.

We are so grateful to you Chris, for sharing your story with us here!

Listen to the full episode:

🎧 Spotify 📺 YouTube

Learn more:

• Chris Shaffer's book about his phage therapy journey
• Chris’ website: Phage Therapy USA
• Case study on Chris’ case, published in the International Journal of Clinical Virology
• Eliava Phage Therapy Center in Tbilisi, Georgia
• 2021 FDA-NIAID Workshop on Phage Therapy transcripts mentioned during the conversation
• Pranav Johri's case study that helped Chris trust phage therapy
• An interview we did with Pranav about his story (YouTube) (Blog post)
• PASTEUR Act, which Chris mentions: a bill in US congress (still has not passed) on changing how antimicrobials can be paid for by the government

Phage Therapy Access and Implementation · Phage Community and Collaboration · Personal Experiences and Reflections

"Why can’t anyone seem to do a positive phase two clinical trial for phages?" This question has echoed through the phage therapy field for over a decade. In our latest Podovirus Podcast episode, Jonathan Solomon, CEO of BiomX, delivers the answer many have been waiting for – his company just completed a successful phase 2 trial for diabetic foot osteomyelitis using phage therapy, showing not just microbial effects but actual clinical improvement with statistical significance.

“We take it for granted that phages are safe. The real question is: can we show convincing clinical data? And now we have.”

With 40% of DFO patients facing amputation, this is an area of massive unmet need. Jonathan shares how their trial was designed (he credits the team at Adaptive Phage Therapeutics, which began the trial prior to its merge with BiomX), and how it achieved what many of us have almost given up on: positive, statistically significant clinical efficacy data for a phage therapy treatment*.

Highlights from our conversation:

• BiomX's diabetic foot osteomyelitis study showed approximately 40% improvement with statistical significance – and in patients where Q-tip could reach bone, 12 out of 13 phage-treated patients showed significant tissue regrowth versus only 5 out of 9 in the placebo group
• Their personalized approach used a single optimized phage for each patient rather than a cocktail – selected from a bank of Staph phages
• The "head of the snake" theory held true – treating only Staph aureus in polymicrobial infections was effective, suggesting not all bacteria present need targeted phages
• The Navy and Defense Health Agency have been significant funders due to concerns about drug-resistant infections in wounded troops from Ukraine, potentially offering continued support for future trials

A few snippets from the episode…

On clinical trial design and working with clinicians:

"We kind of talked to everyone and it kind of goes to the question of dynamic range, like what can we expect to see? And the experts told us, look, they are being treated with antibiotics. So in 12 weeks, you should see a reduction of 40% in the size of the ulcer on the placebo group... So then you kind of say, okay, well, if placebo is shrinking by 40%, what do you guys want my phage treatment to be? Is it 5%? Is it 10%? And they're like, no, it's got to be 20 to 30% because if you don't do anything beyond that, then it's not really moving the needle."

On their phage dosing and delivery approach:

"So again, I think this is all credit to the folks at Adaptive Phage. And I think the rationale was to follow compassionate use cases... I think the rationale was a bit like oncology — a single dose for debulking bacteria [using] the IV route. And then topical is trying to break down biofilm, and that would take a longer time to have the effect."

"We've insisted on taking bone biopsy and growing bacteria out of it and making sure that the phage is [effective against strains] in the bone, so very careful screening criteria, long dosing and quite a complicated regimen... Because we've insisted on taking bone biopsy and culturing bacteria, there was a major dropout because not all the centers know how to do a bone culture or bone biopsy and grow bacteria out of it successfully."

On the commercial potential and investor considerations:

"Just think about the numbers. There are 160,000 lower limb amputations in patients with diabetes alone. 85% of them are driven by DFO, right? So just think about the number of amputations... if you have DFO, the patient will end up 40% of the time with an amputation. By the way, the five-year survival after an amputation is like 50%, right? That means it's worse than cancer. Like it's worse than the worst cancer."

"The typical pharma executive will say, ‘look, infectious disease is not that exciting. And phage is this new modality. I mean, it's not even a new modality, right? You guys have been talking for the last hundred years, right? Is there convincing data? Because some people would say it just doesn't work, right? People have been talking for such a long time. It just doesn't work. Let's move on.’ And I think hopefully now, right, we put a dent in that theory."

On biofilm research needs:

"The thing that has sort of surprised me is that there isn't that much work in biofilms. I think there's a lot to learn about biofilms. In DFO, you can literally image the biofilm... I suspect that there's so much more work that we can do in biofilm to harmonize assays and understand. And sometimes biofilm could be polymicrobial. So what does it mean?... I suspect that that's one of the leading causes for failure of standard therapy"

What's next for BiomX...

Looking ahead, Jonathan envisions a larger pivotal study for diabetic foot osteomyelitis, potentially qualifying for orphan designation and breakthrough status. They're also advancing a phase 2B cystic fibrosis study with 60 patients treated for two months. With strong investor backing from Deerfield and continued interest from the Defense Health Agency, BiomX aims to have one or two approved drugs within five years and to become a "go-to modality" within 15 years.

Learn more!

1. BiomX Positive Phase 2 Diabetic Foot Osteomyelitis Results
2. The details of BiomX/Adaptive Phage Therapeutics’ DANCE trial
3. Recent key opinion leader discussion of Diabetic foot osteomyelitis results
4. BiomX Positive Part 2 Phase1b/2a trial of Cystic Fibrosis results
5. BiomX (NYSE: PHGE) website

Added context

Timestamp ~22:50: Technophage (a biotech company in Portugal) did run a prior Phase 1/2a trial for diabetic foot infections. While the study showed their TP-102 phage cocktail was well tolerated and safe, and reported improved outcomes, the authors report it was underpowered to determine the superiority of TP-102 over placebo, as it ended before reaching the final target of 18 patients due to slow recruitment. Technophage is currently running a larger version of this study.

Of note, there has been one other successful phase 2 phage trial: the Wright et al. ear infection trial published in 2009. This clinical program would have moved to phase 3, but corporate priorities shifted around the time of the ‘08 crash, and it was dropped (check out this interview with trial lead David Harper to learn more). Phage companies have been trying to re-reach this milestone ever since. Finally, it’s been reached.

Want more stories from behind the front lines of phage development?

Subscribe to the Podovirus Podcast!

Spotify: https://open.spotify.com/episode/21MzBWdGUKrm9ujTKj7ist

YouTube: https://www.youtube.com/watch?v=RKF4FgSFfjI

Apple Podcasts: https://podcasts.apple.com/us/podcast/podovirus/id1798745994

Clinical Trials and Case Studies · Phage Therapy Applications · Intellectual Property and Commercialization

Hi everyone,

Hope y’all are doing well and flinging right through spring.

Jessica’s been doing well at Paul Bollyky’s lab, continuing her work on filamentous phages, peeking into the world of phage display, and exploring cell and gene therapy. Hit her up if you’re also going to the ASGCT conference in NOLA in May!

As for me, in case y’all haven’t heard, I’ve been stepping outside the phage, bioinformatics, and microbiology world, to work at an AI company called Groq. Groq provides a way for developers to easily build AI-features like text-to-speech, audio transcription, and summarization. Hopefully, over the next few months, I’ll have time to release some free tools for the research community, like multi-lingual Pubmed search, so stay tuned!

We’ll also be exploring some more short-form and AI-assisted article formats, as we’re growing shorter and shorter on free time!

But for the moment, Phage Directory continues to live on as a blog, and Jessica has some interesting papers she’s been adding to the pipeline.

~ Jan

For this month, here’s what we published:

Our Phage Picks for April 2025

by Jessica Sacher, Jan Zheng

Continuing our favorite format, Jess picked her most-used phage labs methods papers. The first paper is her favorite phage production protocol, “Phage on Tap,” which prepares high-titer phage stocks within two days (we used this in Australia). The second paper is a qPCR-based phage quantification method. As a bonus, I’ve been reading the excellent biotech/AI blog OwlPosting, and found a curious post about optogenetics, and how it’s always been promising, but never really took off (sounds familiar?)

Does every scientist need an AI co-scientist?

by Jessica Sacher

In this Podovirus podcast recap, Jessica shares highlights from a conversation about Google’s new “AI co-scientist” tool (still in beta testing). Two Imperial College London phage researchers José Penadés and Tiago Costa have been the first to test this tool — they gave it a bunch of data surrounding a complex phage biology question (regarding PICIs — a type of phage parasite) that took them years to explain. The AI suggested the same key experiments they’d eventually landed on after years of chasing other threads, which made them all contemplate the role of AI in science. While I think AI-based experimental “auto-suggestion” is incredibly useful (and getting better by the month) — the real scientific work is still being achieved in the lab, the old-school way.

Tips from the lab: Rapid lab hacks

by Jessica Sacher

Here’s a new format! ‘Tips from the lab’ intends to be a new format for sharing concise yet effective ideas for getting the lab in order. This is the stuff people learn the hard way, but that don’t make it into protocols. In the first issue, Jess covered: how to improve phage purification, how to prevent cross-contamination, chromatography techniques for filamentous phages, and tricks for checking in on plaque assays and boosting phage production. Oh and if you have your own hacks, we want to share them!! Email us!

Lessons from viral vectors, CAR-T, & RNA

Throwback: In light of Jessica going to the cell & gene therapy conference, here’s a good throwback that fits the subject!

~ Jan & Jessica

Phage Production and Manufacturing · Technical Aspects of Phage Research · TechBio

Hello everyone!

It’s Phage Picks time! The theme of this week is ‘bread and butter’ — what are some bread and butter techniques for your phage lab, that you can rapidly adopt to dial in your quality/ease of work? The two papers I’m picking today surround solid yet simple methods for two major foundational things you will likely do repeatedly in a phage lab — producing/purifying phages, and quantifying/quality checking them.

One is my favourite methods paper of all time — Phage on Tap. I had to double check I hadn’t already chosen it as a Phage Pick! This paper is my favourite go-to paper for purifying any given phage. It’s lab-scale (small cultures), but high-purity (high enough to go into patients if you needed it to). The octanol method for removing endotoxin is still one of my favourite methods — reliable and extremely low-cost (you don’t need special equipment).

The other paper is one I haven’t personally tried, but I find myself telling people about all the time (and constantly telling myself to do this for all my favourite phages). Essentially it covers qPCR of phages, relating that to plaque assays, and monitoring that ratio for each phage as a way to have a holistic view of its quality/freshness/health. I think all phage labs should ultimately be striving to have PCR primers for their main phages, be checking titres this way (as well as through plaque assays), and using this combo of methods as a periodic quality/integrity check for their phages. If qPCR shows higher phage numbers than plaque assays, maybe you’ve got a degrading/less-than-fresh sample. If they’re staying constant, maybe you’ve got a happy, extra-stable phage! I’m not saying I currently do this, but I aspire to, and this paper is never far from my brain.

Also, beyond qPCR, even regular PCR would be a fabulous addition to routine phage quality checking, and I think we forget that this is so readily available to us. Instead of resequencing phages when we suspect that genetic drift is happening, or that a contaminant has been introduced, or when we’re not sure if two plaque morphologies in the same sample are two different phages or not, we could just be doing a PCR test to verify we get the band we should get for that phage. It’s not everything, but it’s something, and once you have the primers, it can be something you routinely check each time you make a new batch of phage.

Anyway, I hope you enjoy these two papers, and/or file them away in PDF form for a rainy day! Also, if there’s a methods paper you like, where you can say ‘I’ve used this and it works’, please email me! I’d love to feature more papers like this. A methods paper that truly works is gold — we should tell each other about them! (And write to the authors and tell them too — getting an email like that from a fellow scientist who actually read and used your paper is also gold).

Lastly, at the end of this issue, we’ve also got a bonus pick from Jan — what is optogenetics, and why hasn’t it changed the world yet?

Phage on tap–a quick and efficient protocol for the preparation of bacteriophage laboratory stocks

What is it about?

This paper describes a method called "Phage on Tap" (PoT) for quickly and efficiently preparing high-titer, homogeneous phage stocks for laboratory use. The method involves phage propagation, purification by filtration and chloroform treatment, concentration by ultrafiltration, removal of endotoxin using octanol, and storage in banks at 4°C. It can produce clean, high-titer phage stocks (up to 10^10-10^11 PFU/mL) within just 2 days, with limited equipment (just standard things most microbiology labs have) compared to 5+ days for traditional methods.

Why I'm excited about it:

1. The PoT method is straightforward, fast, and efficient, allowing researchers to generate high-titre, homogeneous phage stocks quickly, without the need for specialized equipment.
2. It includes steps for endotoxin removal, which is important for using phages in eukaryotic systems (from cell culture and immunology experiments to human patients).
3. The method is extremely well-laid out/described, and it works almost every time I’ve used it, across many phages!

Overall, this protocol addresses key challenges in preparing phage stocks for research, is very well adapted to researchers in academia, and extremely clearly communicated. It could/should become a standard, default method (at least to start with) for phage biologists.

Source: Bonilla, N., Rojas, M. I., Cruz, G. N. F., Hung, S. H., Rohwer, F., & Barr, J. J. (2016). Phage on tap–a quick and efficient protocol for the preparation of bacteriophage laboratory stocks. PeerJ, 4, e2261. DOI: 10.7717/peerj.2261

Development of a qPCR platform for quantification of the five bacteriophages within bacteriophage cocktail 2 (BFC2)

What is it about?

This study developed a method using quantitative PCR (qPCR) to quickly and accurately count the number of phages in a cocktail of 5 different phages called BFC2. For each phage, they compared qPCR counts to the counts they got using plaque assays, the phage field’s gold standard for phage counting.

Key findings:

• The qPCR method could accurately detect and count all 5 phages in the BFC2 mixture
• qPCR was faster (3 hours vs 2 days) and less work than plaque assays
• As expected, qPCR gave higher phage counts than plaque assays (since it detects all phage DNA, not just active phages)
• The ratio between qPCR and culture counts was consistent for each phage type (!)
• qPCR was more reproducible between different operators than the culture method

Why I'm excited about it:

This paper takes phage quantification from 1-dimensional to 2-dimensional. Using qPCR for phage quantification is not a novel concept in itself, but they describe doing this systematically for a set of phages, and specifically looking at the ratio for the two across the phages.

Being able to quickly determine phage concentrations is important for quality control in making therapeutic phage preps, quality control during phage research projects in general, and for monitoring phage levels during phage therapy. And bringing in a second method (and specifically looking at the relationship between the two methods) gives a fuller picture of not only phage quantity, but phage quality.

Also, the fact that the ratio looks to be phage-specific was really cool to me — this suggests you can use this as a fingerprint/a rapid check to confirm the identity/freshness/stability of your phage. Ratio off? Maybe the phage isn’t behaving like it should — maybe it’s not a fresh batch. Or maybe a contaminant was introduced?

While the plaque assay method is still needed to check if phages can infect bacteria strains (and is the only readily accessible, reliable method for counting the ACTIVE phages in your solution — at least under the conditions you’re testing), qPCR offers big advantages for rapid quantification of phage stocks. It makes a ton of sense as a complement to plaque assays, and should be used routinely to monitor stocks.

P.S. When I was at Phage Australia, I wrote a couple of blog posts on phage qPCR to monitor phage levels during phage therapy: part I and part II.

Source: Duyvejonck, H., Merabishvili, M., Pirnay, JP. et al. Development of a qPCR platform for quantification of the five bacteriophages within bacteriophage cocktail 2 (BFC2). Sci Rep 9, 13893 (2019). https://doi.org/10.1038/s41598-019-50461-0

— Jess

Bonus: Optogenetics could change the world. So why hasn’t it?

This essay from a Stanford undergraduate student covers optogenetics, what it promised for many years, and why it never took off as everyone hoped it would. Optogenetics is basically a way to use light to control bacteria, similar to how our human eyes respond to light with photoreceptors, converting light into chemical signals.

Optogenetics has been a promising avenue for years (for example in neuroscience and medical treatments), but because of continued difficulties in gene therapy (and delivering light to specific body areas), it’s never reached its promise.

This article covers the challenges, road blocks, and explores whether it’s even worth exploring. It’s an inspiring Sunday afternoon read!

Article: https://www.owlposting.com/p/optogenetics-could-change-the-world

— Jan

Phage Production and Manufacturing · Technical Aspects of Phage Research

This week, we’ve got a great story for you — both from a phage biology perspective, and a ‘should we use AI for science?’ perspective. This is a story that was covered last month by Forbes, BBC, and several other news outlets, mainly because of how exciting it is for Google, who created a new AI tool for scientists, and for everyone who cares about AI/wonders if it’s worth the hype.

I was super excited when I saw this story (because I am always excited if we scientists get new tools to help us go faster), but also because this mainstream news-worthy story was actually involving a researcher I knew and had followed for years! It was a phage story! So I had to reach out to José and see if he’d let us interview him on Podovirus.

Excitingly, even though he’d been exhaustedly fielding media calls all month, he was happy to have a chance to talk to our community of phage scientists, to get into the nitty gritty of the science, and discuss the reality, his own skepticism, the humbling experience of working with an AI who saw something he didn’t, and what it all could mean for science. He and his colleague Tiago have published all of this, from the biology to experiment they did to test the AI tool (links below!), but first, here’s their candid conversation (and here on YouTube) — we hope you enjoy!

— Jessica

When Imperial College London professor José Penadés and his lab discovered that phage satellite entities called capsid-forming PICIs (Phage-Inducible Chromosomal Islands) could swap tails to infect different bacterial species, his lab had solved a puzzle that had eluded them for years.

Just before he was ready to publish, however, his collaborator Tiago Costa (another Imperial College Professor) told him about a new AI tool (Google’s ‘AI co-scientist’) he was testing. Tiago was really impressed with the hypotheses it was generating, and this piqued José’s interest.

The two developed a plan. Instead of waiting years to prove whether the AI was right or wrong about the experiments it was proposing for Tiago’s project, could they test the system in reverse? Could they instead first show the AI all of José’s lab’s unpublished data about PICIs, which they’d only finally figured out how to interpret, and see if it could get the same answer?

To their surprise, delight (and no doubt a bit of dismay), the AI quickly came up with the key experiments it had taken them years to consider doing.

Why? According to José and Tiago, it has to do with their own scientific biases, which made them not contemplate some ‘obvious’ explanations for the patterns they were seeing.

On the newest episode of Podovirus Podcast, Joe and I spoke with José and Tiago about the entire process — from the fascinating PICI biology question they were mystified by (and finally solved), to the experiment they did to test the AI, how it humbled and inspired them, and how they now think about how to use AI for science. Lastly, we dig into what this all means for us human scientists. What scientific biases might we each be bringing to the table? When is ‘beginner’s mind’ exactly what we need to make sense of our data? Can we use AI to help us find the patterns in our work that might be lying in plain sight?

Highlights from our conversation:

• José's team discovered that capsid-forming phage-inducible chromosomal islands (cf-PICIs) can exist as "tailless capsids"; and that—contrary to prevailing assumptions in phage biology—these tailless cf-PICIs acquire tails after assembly and cell lysis by binding to free tails from different phages
• This modular "mix and match" capability allows PICIs to dramatically expand their host range by pairing with tails that target different bacterial species or genera
• When presented with the same question, a bunch of unpublished observations, but none of the conclusions, Google's AI Co-Scientist tool suggested the correct mechanism (along with a set of testable hypotheses to prove it)
• The researchers estimate that using this AI tool sooner could have reduced their failed experiments (and unneeded rabbit holes) from ~90% to 50%, potentially saving years of work
• The discovery itself has significant implications for phage therapy, as it suggests a new way synthetic phages with customizable and much broader host ranges could be engineered. Finally, after years of work, they have the mechanism (and now they also have an AI tool to hopefully help them crack their next biological mystery a lot faster!)

A few snippets from the episode…

On how scientific bias delayed their discovery:

José: "We were biased. That's the problem we had. For many years, I always thought—and all phage biology people think—that after infection, what you have are infective particles with the capsid and the tail. We didn't understand why we had PICIs that could be induced but didn't get transferred... We were so biased we couldn't see what was actually happening."

"We have been thinking a lot, okay, how this machine got the right answer. Okay. And it's very frustrating. It's very frustrating because we have the answer there and we didn't see it.”

On the AI's unbiased approach:

José: "Google [AI co-scientist] didn't understand what's happening here, but made the most simple connection. If you are in many different species, and to go to these species you need a tail, you are binding to different tails. That's it... We all knew that the tail determines tropism. It was obvious, but someone needed to tell us."

On adapters and connectors determining specificity:

Tiago: "The tropism will be always determined by the tip of the tail... But the neck region, where the port or the adapter and the connectors are, will have some role in determining that tropism... by generating a promiscuous or a very specific structure that binds only one or several phage tails."

What's next:

José and Tiago are now investigating whether regular phages (not just PICIs) can also exchange tails in natural environments. They're developing IP around synthetic phage particles with customizable host ranges, which could potentially overcome one of phage therapy's biggest limitations—narrow host specificity. Additionally, their labs are continuing to work with Google to test the co-scientist on new research questions, comparing AI-driven hypotheses with experimental results.

Listen to (or watch!) the full episode:

• Watch on YouTube
• Listen on Spotify
• Listen on Apple Podcasts

Learn more:

• Google's Co-Scientist preprint - Details on how the AI system generates and ranks competing hypotheses
• Penadés & Costa's preprint on PICI mobility - The research paper detailing the novel mechanism of PICI transfer
• The team's evaluation of Co-Scientist - Their analysis of the AI system's performance on phage biology questions (and comparison to other AI systems like ChatGPT)
• Want to learn more about phage satellites? Check out this new Nature Reviews Microbiology by José's team
• José's lab: https://profiles.imperial.ac.uk/j.penades/about
• Tiago's lab: https://profiles.imperial.ac.uk/t.costa

Phage Biology and Research · TechBio

I thought I’d do something new this week — maybe it’ll become a recurring format for C&T issues if people like it! I’m going to share some off-the-cuff tips and tricks that I find myself saying to labmates who come to me asking for phage help. I would love for readers to submit theirs, and I’ll compile them and share them here! Send your name or choose to be anonymous, it doesn’t matter to me!

Tip #1: Clean & verify

Are you purifying phages using a reusable system, like an AKTA Pure, AKTA Flux, or a reused PD-10 or Endotrap column? AFTER you clean it (with whatever cleaning reagent - 1M NaOH is my go-to, assuming the column/system is compatible with it), run some buffer through it, and do a phage titre on that buffer before you use the column/system for another phage! If the titre shows even ONE plaque, I would repeat the cleaning procedure, and repeat the titre until you get zero (or at least below a limit of detection you’re comfortable with). Otherwise, next time you use it, you’ll be cross-contaminating your own samples (which is even an issue if you’re using the column only for versions of the same phage… technically, every new batch of phages is different from any other batch, due to genetic drift potential). Clean your equipment, then make sure it’s clean (by titering what flows through it) before reuse!

Tip #2: Try steric exclusion chromatography for filamentous phages

Purifying a filamentous phage, and feeling like you’re not getting rid of enough endotoxin? Have you tried hydrophobic interaction chromatography, but your phage isn’t really binding the column? Don’t have time (yet) to engineer an affinity tag for purification that way? Try steric exclusion chromatography, where you mix phage with a PEG-based buffer, and run it at high speed through an AKTA Pure-type protein purification system. Use this paper as a reference - their method is working really well for me! (Thanks to Arne Echterhof for directing me to this paper - I am so glad to have it!) I will try it on lytic phages next, and see if it works similarly.

Tip #3: Change your top agar percentage to expose invisible phages (or shrink overly huge plaques)

Are you not seeing plaques in your plaque assays, but think you should? Consider changing your top agar percentage. If you are using 0.5% (generally a standard recipe), you could try making a batch of 0.4 or 0.3 %, and re-plating your samples. Really large phages sometimes don’t diffuse well through agar, so the plaque will be so tiny you can’t see it. It can help to lower the agar percentage so more diffusion happens, and plaques become more visible. Alternatively, if you’re getting annoyed that your plaques are too big, and hard to count, you can go higher, say to 0.6 or 0.7% agar. This may shrink them so they’re easier to count. Of course, there’s a limit — too little agar will not solidify (I think 0.3% is the minimum I’ve been happy with?) and too much will probably block diffusion too (but I haven’t really pushed it in that direction - let me know if you do!).

Tip #4: Plaque size so big you can’t count? Check plaque assays earlier (~after 2 hours instead of overnight)

Related to the above - are you working with phages that make giant plaques, way too quickly, like phiX174 phages? You might try incubating them only for 2 hours, then checking them up next to a window to see countable plaques. I was shocked that this worked! It also gives you a convenient ‘same-day’ titre, which is so luxurious! Instant results! Alternatively if you don’t have two hours because you’re about to leave, you can leave it at room temperature overnight instead of 37C. This will help slow things down and give you a fighting chance to count them before they crowd into each other. Of course, you could always just do more dilutions so you do have countable numbers! Also, this will only work with sufficiently fast-growing bacteria (for me it works for E. coli but not for Pseudomonas).

Tip #5: Titres too low? Give broth more surface area (and add calcium and magnesium!)

Are you sad that your phage titres are too low? You could try a few things that sometimes work, like using an empty Petri dish and using that instead of a flask of media. I have found 10 mL in the bottom of a Petri dish (no agar layer! And no shaking!) led to at least a log higher phage production with some phages, at least Campylobacter phages. I think because of the high surface area to volume ratio - I think some cells/phages like that. Alternatively, my other go-to is to add 10 mM CaCl2 and 10mM MgSO4 to your broth media - phages love these cations! Adding them to existing lysate can also make your titre appear much higher - maybe because it helps phages not clump together so much?

There you have it - those are the 5 tips that came to mind this week. Let me know if any of them help you, or if they don’t work at all, or if they spark ideas of stuff you’ve tried that could help other community members! (Write them down and email them to me now, before you forget!! Rough is fine - I can fix sentence structure if you send me rough bullets).

Until next time,

Jessica

Phage Production and Manufacturing · Technical Aspects of Phage Research · Lab notes

Hi Phans!

We hope you’ve had a great March! Jessica and I just moved to Palo Alto, to be much closer to her Stanford lab. We’re now just five minutes away!

The Cover image pictured above is of the Stanford Dish, a radio telescope built in 1961 by the Stanford Research Institute, and is on Stanford campus. They still do research there, but nowadays it’s a popular hiking trail called “The Dish!”

Meanwhile, here’s what we covered this month in Capsid.

Best,

~ Jan & Jessica

A tale of three viruses (and how I discovered them)

by Aaryan Harshith

In this article, Aaryan Harshith recounts his year-long journey as an undergraduate researcher at Stanford's Bollyky Lab, where he successfully isolated and characterized three novel bacteriophages. Starting with wastewater samples from Stanford's Resource Recovery Center, Aaryan details his process of phage isolation, purification through multiple rounds of plaque assays, genome sequencing, and finally electron microscopy visualization. After months of trial and error, his persistence paid off when VIRIDIC analysis confirmed his phages (named Cardinal, Vanta, and Luminis) were indeed new species, with two showing activity against clinical antibiotic-resistant Pseudomonas strains.

Our Phage Picks for March 2025

by Jessica and Jan Zheng

In this article, Jess and I continue to pick our favorite papers from the phage field. Jess picked two papers: one about a hybrid T4-AAV viral vector that can deliver massive DNA payloads (up to 170kb) into human cells, and another on cell-free synthesis of bacteriophages including complex ones like T4. I picked taxMyPhage — a new automated tool from Andy Millard’s lab for classifying dsDNA phage genomes with over 96% accuracy that syncs with ICTV's database.

If antibiotics companies fail, why won’t phage companies?

by Jessica Sacher

In this article / podcast, Jessica interviews Amanda and Mayukh of Phiogen, a company spun off from TAILOR Labs’ efforts to scale up personalized phage therapy. They explore how Phiogen is differentiating itself with their rapid phage evolution tech (the “Directed Evolution Machine”). They also consider new business and pricing model precedents like the recent Live Biologic Products approvals for C. diff infections — which could open the door to $10k per treatment for phage therapy!

Getting phage research funded: Joe Campbell reflects on a career at NIAID

by Jessica Sacher

C&T Throwback: Here’s a fairly recent article covering how to get funding from NIAID, and how to work with Program Officers. With all the funding cuts, this almost feels like a throwback to a previous era.

Phage Biology and Research · Phage Production and Manufacturing · Intellectual Property and Commercialization

Note from Jess: This week I’m excited to cross-post an article originally published by Aaryan Harshith, a Stanford undergraduate student in the Bollyky lab, on his own blog, 'Re: Life'. This article blew me away — I went into it thinking it was a technical blog post about techniques, and was suddenly captivated by the narrative (instead of talking about the concentration of top agar in a plaque assay, it whisks us away early with ’A murder in France’).

Please enjoy Aaryan’s article about his search for phages this past year, his thoughts on science, and just the joy of reading his writing, which reads like a novel. And then go subscribe to his blog, ‘Re: life’ for more thoughtful commentary and consideration about this whole scientific enterprise we’re all part of!

-- Jess

Note: In this article, I retrace more than a year of my microbiological research, and how it led to the discovery of three new species of viruses (1, 2, 3).

This post was written for citizen scientists and the public. As lengthy as it is, much of it still oversimplifies the nuanced, rapidly-evolving landscape of bacteriophage research.

If you’re a fellow phage researcher, stay tuned for a series of posts where I open-source my protocols and illustrate them in technical detail. Or, feel free to reach me directly at aaryanh@stanford.edu.

Table of Contents

1. Exposition

2. A Murder In France

3. The Beginning of My Adventure

4. The Search for Samples

5. Phishing for Phages

6. Isolating Phages

7. Sequencing Phage Genomes

8. The Moment I Discovered a Phage

9. How to Name Your Phage + Other Side Quests

10. Aftermath

11. Acknowledgments

    ---

Chapter 0: Exposition

As I slipped on my blue latex gloves in the throbbing California heat, I couldn’t believe that the friendly-looking man I had just emailed yesterday was about to hand me a tube full of sewage, or that the quaint, retrofitted shipping container sitting in front of me was his office.

Before I could reflect on my situation, I was holding a red, plastic bag emblazoned with biohazard logos. On the bike back to lab, I pedaled faster, trying my hardest not to picture the tube warming slowly in my bag, or its unknown contents sloshing around with every pebble and turn. Searching for some moral support, I try craning my neck to the sun, but it only seems to get hotter.

In moments like these, it’s hard to keep the same question from coming to mind: “How did I get here?”

Chapter 1: A Murder in France

Our story begins more than a hundred years ago, in the bustling city of Paris.

Crouched over his workbench, an eccentric scientist by the name of Félix D’Herelle is busy investigating the murder of his bacteria. The evidence in his hands was undeniable—a nutrient plate that should be flooded dysentery colonies, instead wiped clean by an invisible culprit. But who—or what—was to blame?

Like any proper biologist, D’Herelle got to work. After pumping the remains of his bacteria through an ultra-fine Chamberland filter, he found that the liquid that flowed through was just as destructive to his colonies as before. With his elegant experiment, D’Herelle unraveled an important clue: whoever the murderer was, it was far too small to be a bacteria, parasite, or any cell for that matter.

But what was it, exactly?

Félix D’Herelle inspecting cultures in his lab. Source

Months later, still in the thick of his confusion, D’Herelle publishes a landmark paper titled “An invisible, antagonistic microbe of the dysentery bacillus”. Two years later, in 1919, the first patient was cured of dysentery with phage therapy. This was the first documented discovery of a bacteriophage, and perhaps the worst timing share it with the world.

With the isolation of penicillin in the 1930s, excitement in phages waned in favor of cheap, reliable, and scalable antibiotics. In the coming decades, life expectancy skyrocketed, and the global population grew as millions of children survived into adulthood. Humanity had finally liberated itself from the tyrannical reign of bacteria, or so it seemed.

In a strange twist of fate, we found ourselves in much the same condition today. Bacteria are back—and they’re better.

A 2018 study from Addo et al. found that nearly 50% of E. Coli strains isolated from water samples were resistant to one more frontline antibiotics. Roughly speaking, this trend holds across the board, and antimicrobial resistance is projected to take more lives than cancer by 2050.

Context on the already massive scale of antimicrobial resistance (AMR). Source.

In retrospect, though, this seems inevitable. Knowing the explosive rates of bacterial replication and mutation, all it takes is a handful of calculation to suggest that bacteria can develop resistance to antibiotics faster they can be designed, tested, put on the shelf.

This is the exact gap that has rekindled phage research around the world. As the natural predators of bacteria, phages satisfy many of the constraints of antibiotics by design. They’re incredibly safe, abundant in nature, and often selective enough to target bacteria with single-strain precision. And unlike antibiotics—which are static molecules that often take years to update— phages have nimbly kept pace with bacterial evolution since the dawn of cellular life.

At a glance, phages seem to embody the perfect antibiotic—that is, assuming we can understand them properly.

Chapter 2: The Beginning of My Journey

Even with its unprecedented comeback, the space of unknowns in phage research remains unfathomably large.

As I write this article, our latest data suggests that humanity has characterized ~15,000 unique phages. For context, we’ve meticulously documented the existence of at least 350,000 species of beetles1.

Save for our fundamental understanding of how phages infect and replicate, the vast majority of phages and hosts haven’t even been described yet. Without a grasp of what phages exist, where they’re metabolized, or how they act, how can we expect to use them in life-or-death situations?

Back when I was a freshman, these were the kinds of open questions that inspired me to join the Bollyky Lab in Stanford’s School of Medicine. This is where I learned just how promising phages are, as well as how enigmatic they still were. After spending nearly two years of investigating our lab’s speciality—phage therapies against Pseudomonas lung infections—I finally had chance to launch a project of my own.

Fortunately, there was an idea that had already been tugging at me. If bacteriophages were the next unexplored frontier in biology, I wanted to help advance it. I wanted to discover a new phage.

Chapter 3: The Search for Samples

One could say Félix left quite the legacy. Fast forward more than century, he’s currently the reason I’m currently riding my bike at full-tilt down Jane Stanford Way.

Which brings me back to my first original question: why was I here?

To understand this, we have to understand what makes viruses special.

In one of his most famous books, the American microbiologist Dickson Despommier once quipped that “successful systems attract parasites”.

Like all viruses, phages are parasites that subsist by entering cells, hijacking their molecular machinery, and turning them into virus-making factories. Since their survival hinges on the presence of an appropriate host, phages tend to congregate around them.

Or, to spin this another way: the simplest way to find a phage is to look for its bacteria of interest.

At the Bollyky Lab, our team focuses on Pseudomonas aeruginosa—an opportunistic bacteria that is already responsible for millions of drug-resistant skin and lung infections around the world.

What makes Pseudomonas particularly interesting are its hideouts. Beyond lingering in hospital rooms, bathroom sinks, and the built human environment, P. Aeruginosa is routinely found in soil, ponds, and wastewater.

With a trail in sight, I decided to spend my first weeks of research as a field biologist.

Left: Me and Bill (Bill and I?). Right: A full array of environmental samples.

In 15mL Falcon tubes, I collected stagnated water from Lake Lagunita, scraped soil samples from outside Medical School, and swabbed rooms at Stanford Hospital. Every day meant a new destination, a fascinating person, or another exotic sample to collect.

And that’s how I eventually found myself here, staring up at a weathered, ivy-covered sign that read *“Stanford University: Codiga Resource Recovery Center”.*From a series of cold emails and asking around the academic grapevine, I learned that if you wanted wastewater, this was your place, and Bill Li was your guy. It certainly didn’t look like it, but the murky water I had just been gifted held exactly phages I needed.Which begs the question: how do you get from wastewater to phages?

Chapter 4: Phishing for Phages

Consider zooming into our tube of wastewater.

With enough magnification, it becomes clear that phages aren’t the only residents. If anything, the depths of this liquid resemble a microbial metropolis, where bacteria and fungi tower like skyscrapers, and every sidewalk and street corner teems with viruses and molecules of all kinds.

This quickly presents a dilemma: in such a bustling carnival of microbes, how do we precisely extract the phages we want?

As I soon discovered from closely studying protocols, the short answer is to use the right bait.

In line with most other viruses, phages can only infect bacteria after binding to molecules on their surface. Since this interaction between the surface of a phage and its target is specific, so is the spectrum of bacteria it can infect.

This property—somewhat uncreatively known as host range—is exactly what microbiologists exploit to select phages.

All of this made sense, at least in theory.

But as I stood in our lab and scanned our seemingly endless walls of reagents, I realized that the task on my hands was on a different dimension from the research I had done. By definition, I was looking for something that had never been found before. There were no fifteen-minute guides to follow or reliable results to expect. For the first time, I found myself probing a question that didn’t have clear answers—every experiment was now an act of creativity.

As unnatural as it felt, the only path forward was to try something.

To pull this off in the lab, I began by spinning down our wastewater in a centrifuge. From reading papers like this one, the logic was clear: through its intense rotation, a centrifuge artificially enhances the action of gravity, allowing the components in our samples to spontaneously arrange themselves by density.

A rough schematic of phage isolation from wastewater.

Naturally, this means cells and macroscopic particles would aggregate at the bottom of our tube, while viruses and small molecules would form a clear liquid on top called the supernatant. For now, I decided to extract the supernatant and set it aside.

Meanwhile, I wanted to prepare the ideal bait for our phages—in this case, a liquid culture of Pseudomonas bacteria that I brewed up some hours earlier.

What do you think happens to our swimming pool of bacteria when we treat it with our supernatant from earlier?

Let’s play this through in our imagination. If our wastewater contained any trace of Pseudomonas phages, they would immediately infect the bacteria in our culture and replicate at breakneck speed. Any other phages in solution wouldn’t recognize the host, and their numbers would remain stable.

In essence, we get a system where our phages of interest multiply explosively—easily overshadowing the remaining viruses that can’t infect our host.

Of course, carrying this out was slightly more brutal than this summary lets on. Till these experiments, I’d only ever read about phage screening in papers from established labs. Optimizing these experiments from scratch was perhaps the bumpiest stretch of this entire project.

From left to right: the same culture of Pseudomonas with increasing concentrations of phage, with the sample on the far right containing plain media. Note the stark color gradient as the phages disrupt bacterial growth.

While the first tests took a day to run, honing them into a workable protocol took three weeks of full-time work. It called for a level of testing, critical thinking, and creativity that I never used before.

That said, I still remember how dramatic the results were when this finally worked.

Sitting in a boxy plastic holder were the two defining vials of my experiment. On the left was a negative control—another culture of Pseudomonas that was already taking on a villainous green after a night of incubation. On the right, the same culture treated with our supernatant had already turned a cloudy yellow—the color of trillions of bacteria gushing out their remains at once.

One thing was unmistakable: we had the phages we wanted, and we had a lot of them.

Chapter 5: Isolating Phages

What I said earlier was a lie. Kind of.

If you were careful, maybe you caught it: the method I described lets us begin with any number of phages and end up with only those that infect our bacteria of interest.

But what if our sample contained more than one phage that could infect Pseudomonas?

Before we can say anything meaningful about a phage—whether it’s sequencing, imaging, or anything in between—we need to be sure we’re working with just one kind. Whatever phages our sample harbors, it needs to be as pure as possible.

This presents an even bigger problem than before: how do you separate a population of phages that already infects the same host?

Perhaps unsurprisingly, a quick review of the literature reveals all kinds of solutions, from centrifuging them aggressively to filtering them through sophisticated columns. Having weighed (and tried) many of these options, my weapon of choice soon became a technique called the plaque assay2.

To understand this, let’s return to our vial of phages. On a good run under the right conditions, just a milliliter of our solution can contain 1,000,000,000 (or 10⁹) particles of infectious phages.

If you dropped these phages on a plate of bacteria and waited overnight, you’d end up with what resembles a gaping crater in a field of bacterial proliferation. These structures are known as plaques, and we commonly use them as a proxy to quantify the number of infectious phages in a sample.

The basic mechanism of a plaque assay with serial dilutions. Source

There’s just one issue. Such a potent mix of phages wouldn’t just form a plaque—it would wipe out every trace of bacteria on our plate.

This is useless for same the reason our eyes are useless in blinding light—whether the source is sun or pen laser, the resulting image is undecipherable because it falls far past the threshold of what our eyes can measure. When overwhelmed with phages from all directions, our bacteria suffer the same fate.

The plaque assay overcomes this barrier elegant way: by simply scaling down our population of phages. Instead of completely annihilating bacteria with our concentrated brew of phages, what if we diluted our viruses?

Straight from the vial, our solution is brimming with phages. But with every successive dilution, our phages become fewer and further between, until they finally produce a countable—and therefore, useful—number of plaques.

For instance, if we know that a thousandfold dilution produces ten plaques, we can infer that the original solution would produce ten thousand.

But that wasn’t my goal. While plaque assay lets us count phages, I began wondering if it might also let me separate them. This follows from the same mechanism discussed earlier. With enough rounds of dilution, the hope is that our phages become so sparse that each and every plaque is the doing of a single phage3.

Oftentimes, this process can even be viewed at the macro-scale, since many phages inherently produce plaques of distinct shapes and sizes. If you then dug out just one of these plaques and re-propagated it in bacteria, the resulting batch of phages would turn out far purer than before.

This didn’t work.

For one, it was clear from the first plaque assay that our sample was home to more than one Pseudomonas phage. After enough dilution, what I once thought was a single, unified plaque soon split into a constellation of smaller ones.

Some of my first plaque assays visualized under a white light (note how the culture on the right is mostly decimated)

This would have been perfect, if not for a glaring issue. When I extracted the phages from a single plaque to run this test again, it still split into a myriad of unique plaques. The message was clear—one round of purification just wasn’t enough to resolve single phages. Much like before, this was a protocol I ran, reflected on, and reworked for weeks.

In the end, the most reliable technique also happened to be the most painful one. I found that the ideal method to arrive at a pure sample of phage was to run the same plaquing and picking technique five times.

In practice, this meant five straight days of boiling agar in the microwave, holding my breath to pick plaques that were nearly invisible, and tending to my phage cultures like I was the parent of a sleep-deprived newborn.

I’m not sure how felt during those moments, but looking back, I laugh—those solitary nights were perhaps the most important part of my research. It taught me what it meant to pour myself into a question until I found an answer. It taught me how to wrestle with the constant ambiguity of my science. It taught me how to be wrong more times than I can count, and still find the resolve to think of my next attempt.

It was confusing, it was frustrating, it was heart-wrenching at times. And it was the best thing that happened to me as a scientist.

Chapter 6: Sequencing Phage Genomes

On the other side of these experiments, I was looking at six vials that I thought contained distinct phages. Now that their contents were finally pure enough, sequencing was the final step left to understanding their identity.

There was just one problem: our lab didn’t have a sequencer. After Googling for alternatives, asking friends and mentors, and sending a dozen more cold emails, I began feeling like a slice of overly burnt toast.

That same day, fueled by a mixture of desperation and anxiety, something finally snapped. Before I could think through the merits of my strategy, I had already zoomed up two flights of stairs, turned a hallway, and gathered the courage to knock on the door of a lab I’d only heard of before.

The scientists I approached that day belonged to the Gao lab, and they later became instrumental in sequencing my phages. As the summer swept by, many of them became my good friends, and our labs still continue to collaborate long after this project has wrapped up.

With the help of Carlo Armijo, one of Dr. Gao’s PhD students, I learned how to prepare, read, and interpret phage genomes with an Illumina sequencer. And in the months that followed, we had the chance to sequence more phages than I can remember.

If you’ve noticed a common theme by now, you might guess that most of these attempts crashed and burned.

An example of a failed sequencing result generated using BLASTn. Note how all the similarity values land above 95% for this phages.

And you would be right. Our very first run, for instance, revealed that our samples still contained fragments of DNA from nearly a dozen different phages. Others produced no reads, or suggested contamination with bacterial DNA. A full two months later, when we thought we had systematically resolved every path to failure, sequencing showed that the phages in our solution already existed.

At this point, I wanted to scream into a pillow.

If this project had any chance of producing novel phages, I would have to start from the beginning. From zero. But as my frustration faded and I reflected on the steps it took to get here, a few explanations began coming into focus.

Perhaps the most glaring issue was my bait. If I was trying to isolate a novel phage—one that no researcher had described before—why was I luring them with the most mainstream strains of Pseudomonas? And if my samples really did contain such a diversity of bacteria, why was I only looking for Pseudomonas phages?

On closer inspection, it wasn’t all too surprising that my experiments had failed.

But I was in far, far too deep to give up now. After taking the weekend off to catch up on lost sleep, I was back, and I was going to set things right.

For one, I snapped on my field gloves and began collecting as many environmental samples as I could. When I wasn’t in the field, I was scouring our storage for the most obscure strains of Pseudomonas I could find. With time, I also expanded my search to E. Coli, which was also supposed to be a common member of the samples I gathered.

More imaging + round two of sample collection

This all extended from the same rationale. I figured that the more phages I packed into a single vial, and the more kinds of bacteria that were available to them, the more likely we were to snag an interesting virus.

All counted, it wasn’t until another two months of experiments till I was ready sequence again.

Strangely enough, that might have been the single upshot to all that drudgery—when my sequences finally arrived unscathed, it felt like I’d conquered the universe. Sometimes, I guess the simple pleasures really are the most touching.

Chapter 7: The Moment I Discovered a New Phage

Throughout my runs with Carlo, we relied exclusively on a technique known as sequencing-by-synthesis (SBS).

In this approach, whole phage genomes are chopped up into thousands of overlapping pieces (“reads”) that are then pieced together computationally. After receiving a series of raw reads from our sequencer and assembling them using an algorithm called SPAdes, we could now read the genetic sequence of our phages as clearly as the text in a Google Doc.

But which of these genomes had something interesting to say? Which of these genomes was entirely new?

For a better intuition, we need to first clarify what we mean by a species of virus.

Viruses are notoriously tricky to organize because they aren’t alive. Whereas a species is classically defined as any group of organisms capable of breeding interbreeding, it isn’t clear how this translates to phages that replicate independently.

A more robust definition of a virus comes from the International Committee on the Taxonomy of Viruses (ICTV), which happens to be the global authority on this exact question. Instead of categorizing a virus on its capacity to breed, the ICTV defines a virus as a new species if its genomic sequence is less than 95% similar (or more than 5% different) from its most closely related existing virus4.

As it turns out, there’s a tool for that. When virologists want to quickly understand the genome of an unknown phage, they turn to a program known as the Virus Intergenomic Distance Calculator (VIRIDIC).

If we want to prove the originality of an essay, we use a plagiarism checker. If we want to prove the originality of phage, we use VIRIDIC. In a nutshell, VIRIDIC accepts the genome of any unknown virus and compares it with those of its closest genetic neighbors. If significantly long stretches of the genomes resemble each other, the program determines the degree of similarity, as well as the likelihood that the resemblance was due to chance.

With multiple phages, VIRIDIC outputs a matrix of genetic similarities—much like the results it produced for the phages we sequenced below:

VIRIDIC analysis for Pseudomonas phages Luminis and Vanta.

It was hard to miss: from left to right, all the genetic similarities between our phages and their closest viral relatives fell below 95%. According to ICTV’s threshold…I had just discovered at least three new species of viruses.

I refreshed the results, and refreshed them again to make sure. A pause, and the same numbers flashed onscreen.

Nearly a year since I set out to isolate a new virus, I had finally accomplished my mission! After exploring what seemed every possible way to mess up, here I was.

Here I was, but what now?

Chapter 8: How to Name Your Phage + Other Side Quests

Apparently, in the niche corridors of microbiology research, it’s standard practice for the discoverer of a virus to choose its name. Several long walks and conversations later, I eventually made up my mind.

I decided that I’d name my E. Coli phage Cardinal, as a friendly homage to Stanford’s mascot. I’d name the other two Vanta and Luminis—partly because they weren’t already taken, and partly because I thought they sounded cool. (I seriously hope I put more thought into the names of my children).

In the weeks that followed, I found myself surging with inspiration—most of which I channeled into understanding these viruses as much as possible.

For starters, I began teaching myself bioinformatics on walks to lab and between bites of dinner. Eventually, this let me run a program called PHANOTATE to computationally predict the identity and function of all the proteins encoded in phage genomes.

As I refined my skills further, I learned how to render my phages on a proteomic tree, which allowed me to map their relationships based on the similarity of their core proteins, as opposed to the sequences of their nucleic acids.

A proteomic tree of phage Luminis. Graphic generated using VipTree

Armed with this arsenal of tools, I could resolve the classification of my phages in granular detail. For instance, I learned that Cardinal belonged to the family Justusliebigviridae. Vanta and Luminis, on the other hand, belonged to the family Yuaviridae, which hardly contained phages to begin with.

Back in the lab, I propagated our phages, meticulously recounted their numbers, and began screening their activity against a broader library of bacteria. This was around the time I realized these viruses were far more interesting than I imagined.

Partway through my analysis, our lab received a nondescript paper bag from Stanford Hospital. Inside was what seemed like a handful of unremarkable Pseudomonas cultures, with the exception that they were multi-drug resistant isolates from patients with severe lung infections.

Phage Vanta plaqued against clinical samples of Pseudomonas (note that the left image uses a different plaque streaking method than the sample on the right)

With the approval of my mentor (and a handful more plaque assays), I revealed that both Vanta and Luminis were active against these bacteria, too. It seemed that the phages I isolated weren’t just theoretical curiosities that infected lab bacteria — they formed plaques in clinical strains of Pseudomonas that were already resistant to frontline antibiotics.

While demonstrating the potential of these phages was exciting, it still fell short of the first time I saw them up close.

Much like bioinformatics, I didn’t begin with a shred of experience in imaging viruses. But through some stroke of magic, I did know of Dr. John Perrino, who regularly ran the electron microscope in our basement. And after pestering him with enough emails, he eventually agreed to teach me how to use it.

I still think back to the day everything came together. After spending the morning staining my phages in uranyl acetate, I was now glued to the microscope, holding my breath while scrolling through magnifications as gently as I could manage.

1,200x

2,000x

4,500x

10,000x

25,000x

50,000x…

And then—almost as if I’d struck the virological lottery—everything snapped into focus all at once. For the first time, I could see the phages I had isolated—I could see their capsids, their stubby necks, and their long, spindly legs. I could see them in the thousands.

Transmission electron micrographs of our phages (note that these samples are not pure)

At that moment, it struck me that I might be the first human to ever see this virus. It felt like I was beholding a world I didn’t have permission to see. Mentally, everything went blank except for a single, abiding thought: *“Ah, I see now—so this is why people become scientists….”*With all the data I’d collected, I figured it was finally time to make these phages official. After organizing and annotating their genomes, I submitted these phages to NCBI, which currently hosts the world’s largest database of viruses.

Last December, I learned that our phages were live—complete with information on their genome, proteins, and metadata. (If you’re curious, feel free to browse through their entries here — 1, 2, 3).

In the meantime, I compiled the rest of our figures into a report and submitted it to ICTV, which will be reviewing our proposal this May. If these submissions get accepted, all three phages will become official canon recognized by scientists around the world.

For the first time in a long time, all that’s left was to wait and watch.

Chapter 9: Aftermath

This article might have been the most ambitious one I’ve ever written. As I finally type out these concluding words, I decided to return to my favorite writing spot on campus—in this case, a plush, red chair overlooking the entrance of Green Library.

Outside the window, I watch as raindrops begins to drizzle the ground, slowly drenching the world in blue. Almost subconsciously, I catch myself thinking that days like these must have been made for reflection. As the rain quietly patters away, I take a moment to recollect my journey.

Last summer—almost a full year ago—I naïvely set out to discover new viruses. Everything else feels like a blur.

The bizarre, strangely nostalgic parking lot where it all began.

I remember biking frantically around campus as a tube of wastewater sloshed around in my backpack. I remember marathons of plaque assays, and the constant, gnawing fear of seeing a lab ghost alone at night. I remember the confusion of my first day in lab, the rush of our first sequences, the joy of seeing my phages for the first time, and everything in between.

It felt unreal, yet oddly satisfying: one year, three viruses, too many memories to count.

For now, though, I’m in limbo—another transit between what happened and what’s coming. What do I want to do next?

Part of me wants to scale everything. If my shoestring experiments could turn out new viruses in a year, what would happen with ten students and a serious budget? What if I launch Stanford’s first viral discovery initiative? If the future of phages matters so much, I didn’t see any reason why we couldn’t pave the way in the years ahead.

The other part of me wants to get further with the knowledge we already have. I want to learn how phages can be engineered, formulated, and directionally evolved maximize their performance. Since phages are so versatile, I also want to learn how they can be applied outside the clinic—in food production, industrial chemistry, materials science, and biodefense.

Somewhere along the way, I begin to realize just how far I’ve come, and just how little I know, and the immensity of all the questions and ideas that I’m not even aware of yet, and…

I’m suddenly shaken awake by the familiar, high-pitched yelp of my alarm. As my thoughts slowly drift back to my surroundings, I realize that I’m in a library and look like I’ve been hit by a baseball bat. Fumbling for my phone, I remember that I have a long day ahead: bio homework, notes to wrap up for tomorrow, then dinner with friends from art history.

I can’t help but smile a little. Research might be over, but my life is still unfolding.

Till the next mission,

Aaryan Harshith

Chapter 10: Acknowledgments

While this research began as an independent project, it now has a small army of collaborators and friends.

First, I want to thank Dr. Paul Bollyky for taking a chance on me and letting me join his lab as a freshman. I’d also like to thank my direct mentor, Dr. Jessica Sacher, who taught me all the techniques I know and always held me to the highest standards as a scientist.

On that note, this research simply wouldn’t have been possible without the generosity of Bill Li from Stanford’s Resource Recovery Center, and Carlo Armijo from the Gao Lab. I’ll always be grateful for your mentorship and support as this project stopped and started more times than I can remember.

Thank you to Stanford’s Existential Risks Initiative for backing my research last summer, and to all my fellow SERI researchers for teaching me about their work between spoonfuls of ice cream.

Thought I’d share an easter egg for those of you who made it so far. This portrait is titled “Me after finishing this article at 3:42AM…”

And of course, thank you to my family, as well as all my friends on campus and beyond whose presence alone kept me sane. Thank you to Ryan Lee, Grace Bedia, Aspen Singh, Iris Qu, Michelle To, Audrey Jung, Jason Chon, Hemal Arora, Dhevin Nandhyala, Melwin Cheng, Sam Yang, Nicholas Tan, and the many, many more incredible Stanford trees who helped me along my way.

Thanks for reading Re: Life!

Subscribe, or you’ll be gobbled up by the flying spaghetti monster.

Source

Footnotes

1 - No shade intended. I make this comparison with a deep respect for the importance of beetles, insects, and the study of entomology.

2 - Among the oldest virological techniques that is still employed in modern labs, originally developed in the 1950s by Renato Dulbecco.

3 - Several microbiologists have pointed out that this assumption is problematic, as differentiating between a true plaque and a larger agglomeration of plaques can prove nearly impossible without prior knowledge of a phage’s unique plaque structure.

4 - While this threshold clearly isn’t ideal (especially since genetic differences don’t necessarily imply outward differences in structure or infectivity), it’s held up surprisingly well against the test of time, and is still convenient enough for the ICTV analyse hundreds of species submissions a year.

Personal Experiences and Reflections · Technical Aspects of Phage Research

Happy Pi day! This week, it’s finally time for a Phage Picks! I think this is the first of 2025? Here are a few of the papers we’ve been thinking about lately, and why.

Jess’ picks:

A prokaryotic-eukaryotic hybrid viral vector for delivery of large cargos of genes and proteins into human cells

What is it about?

One question I’ve been thinking about lately — could phages be useful for delivering genes to eukaryotic cells? Why would we want to use phages for this? As I’ve dug in, it does seem that AAVs and other standard viral vectors have limitations — so there are some genetic diseases that aren’t being solved because the genes can’t be delivered.

As I’ve come to be immersed in since joining the Bollyky lab, phages are actually decent candidates for doing this. They are engineerable, can theoretically take in lots of extra DNA (in part because they tend to be way bigger than human viruses), and so far aren’t seeming that immunogenic. I’ve also heard from some ‘viral vector people’ I’ve been meeting around Stanford that viral vectors are expensive to produce (which makes sense — I assume you have to culture human/animal cells to make them..?). So because of this, it’s hard and expensive to get high enough doses to treat people. (I gave a presentation on phages to a biotech entrepreneurship program I’m going through now (Spark) — more on this later — and people have been saying ‘wow you can get 10^9 pfu/ml? 10^11? So easily? That’s so high, nice!’. So this seems like another area phages could compete — much cheaper and easier to make lots of them.

With that intro, here’s a cool paper I found. Instead of using phages directly, it physically links T4 phage to AAVs (adeno-associated virus; one of the workhorses of delivering genes to eukaryotic cells for gene therapy).

Why would they do this? AAVs can only accept ~5kb of DNA! But T4 is massive… it can contain hundreds of kb! But T4 doesn’t get into eukaryotic cells very efficiently. And AAVs are pretty good at that.

How did they do this? With good old biotin-streptavidin linking, via T4’s decorative head proteins (Soc and Hoc). AAV is tiny by comparison, so it’s basically T4 with AAVs decorating its head.

Did it work? Yes! They delivered up to 170 kb into eukaryotic cells — apparently the largest payload of foreign DNA delivered to date! They also showed they could deliver an array of different payloads at once; they put luciferase DNA in the T4 genome, GFP in the AAV genome, AND displayed β-galactosidase on T4’s surface. All delivered all at once to the same cell. The efficiency was 40,000 fold higher when the AAV was attached, compared to regular T4. And it worked biologically too — they wanted to see it protect against flu and plague (both at once!), and it did that in mice.

Overall this is a pretty cool proof of concept that we can deliver large payloads to eukaryotic cells with the help of phages, both on the DNA side and on the protein side. This means we can unlock treating more genetic diseases than just the ones that happen to have a short region (or max a few genes) to correct. And it also means we could make more multivalent vaccines that protect against many antigens at once.

I’m going to be following this field (and working on it in the lab too — more on this soon!), and would love to hear from anyone playing with phage-mediated gene (or peptide) delivery to human cells! Would love to read any cool papers you’ve found!

Paper: https://www.science.org/doi/10.1126/sciadv.aax0064 Zhu, J., Tao, P., Mahalingam, M., Sha, J., Kilgore, P., Chopra, A. K., & Rao, V. (2019). A prokaryotic-eukaryotic hybrid viral vector for delivery of large cargos of genes and proteins into human cells. Science Advances, 5(8), eaax0064.

Cell-free TXTL synthesis of infectious bacteriophage T4 in a single test tube reaction

What is it about?

I’ve been producing and purifying phages for many years now; in my PhD I made Campylobacter phages so I could study their biology. In my postdoc at Phage Australia I made a potpourri of phages against whatever pathogen was infecting the patient we were aiming to help that month.

Now I find myself in a phage engineering lab, thinking about using a small defined set of phages for delivering genes and peptides to the body. My default has been to use the same methods: grow bacteria, add phage. Tweak the conditions until you hit a reasonable titre at the end. Spend most of your time purifying out the bacterial crud so you don’t confound the results of your experiments / cause any unwanted immune reactions / hurt anyone, and proving that the ‘crud’ is gone (endless endotoxin assays).

But now that I’m moving toward only working with a small set of phages (less than 5), what if I changed my approach? With just a few phages you can develop targeted purification strategies (like adding a his tag and using affinity chromatography). But also, what about cell-free approaches, so we don’t have to deal with bacterial membranes and cell debris at all? Where is this field at?

In my digging I came across this paper — it’s from 2018; they showed that T4 phage could be produced in vitro — a one-pot, cell free transcription-translation system called ‘TXTL’. This was a big deal because to that point, only smaller, simpler phages (MS2, φX174, and T7) had been made in vitro without cells. To make T4, you need 1,500 proteins to be self assembled — pretty insane! One thing they talk about in this paper is the importance of PEG to cause ‘molecular crowding’ (pushing everything closer together), which they think approximates cellular compartmentalization. When they added PEG they saw a 100,000 fold increase in efficiency.

I also wanted to follow up on what’s been done since then, to get a sense of where cell free phage making is at. This led me to this 2022 paper, authored by Quirin Emslander and Kilian Vogele (who Jan and I actually got to meet back in 2018 when we visited Jean-Paul Pirnay’s group in Belgium; they happened to be visiting at the same time!). In their paper they show cell-free synthesis of phages against a bunch of pathogens. They also used mass spectrometry to follow the cell-free synthesis and compare it to normal phage propagation (and saw for T7 that the early-middle-late expression that’s been documented for all these decades actually holds true in the cell free system! The process happens slower though — probably nutrient-limited).

Of note, they could detect more phage proteins via mass spec than previously done with normal propagation (meaning they validated the existence of a bunch of hypothetical T7 proteins!). And it makes sense — this would be easier without the cell proteins in the way.

Overall, to me this second paper highlights how working in these cell free systems can generate a ton of new biological insights too (previously I was only thinking of it from the applied, ‘how can we make more clean phages faster’ sense).

On a final note, Kilian and Quirin are working to commercialize the cell-free system through their Munich-based startup, Invitris. Very excited to see where this takes them! Kilian and Quirin, if you’re reading this (do you still subscribe after all these years?), hello! Let us know how you’re doing, and what’s next in the world of cell-free phage production!

Paper 1: https://academic.oup.com/synbio/article/3/1/ysy002/4821891 Rustad, M., Eastlund, A., Jardine, P., & Noireaux, V. (2018). Cell-free TXTL synthesis of infectious bacteriophage T4 in a single test tube reaction. Synthetic Biology, 3(1), ysy002.

Paper 2: https://www.sciencedirect.com/science/article/pii/S2451945622002355?via%3Dihub

Emslander, Q., Vogele, K., Braun, P., Stender, J., Willy, C., Joppich, M., Hammerl, J. A., Abele, M., Meng, C., Pichlmair, A., Ludwig, C., Bugert, J. J., Simmel, F. C., & Westmeyer, G. G. (2022). Cell-free production of personalized therapeutic phages targeting multidrug-resistant bacteria. Cell Chemical Biology, 29(9), 1434-1445.e7.

Jan’s pick:

taxMyPhage: Automated Taxonomy of dsDNA Phage Genomes at the Genus and Species Level

What is it about?

taxMyPhage is a new automated tool for classifying dsDNA phage genomes at the genus and species levels. They built a workflow to process a MASH database of phage genomes classified by ICTV with BLASTn. The database will be synced with ICTV’s Virus Metadata Resource, which contains all classified virus genomes — which means the classifications will be able to stay up to date.

Why I'm excited about it:

This paper is exciting (despite it being tax season in the US) because phage genome classification has always been one of those tasks you wished could just magically get done — kind of like washing dishes or doing actual taxes. Well, taxMyPhage takes us one step closer to that reality, with an automated phage classification workflow at the genus/species levels (with 96%+ accuracy to boot!).

Because it’s available both from their website and as a CLI tool, and because they’ve standardized the workflows, it’ll save phage labs many hours a year of tedious bioinformatics work. Thanks Andy, Thomas, and team!

(Also, any project that moves any projects from R to Python is a win in my books. The tool is also free and open source!)

~ Jan

Tool: https://ptax.ku.dk/

Paper: https://www.liebertpub.com/doi/10.1089/phage.2024.0050

Source Code: https://github.com/amillard/tax_myPHAGE

Millard, A., Denise, R., Lestido, M., Thomas, M. T., Webster, D., Turner, D., & Sicheritz-Pontén, T. (2025). taxMyPhage: Automated Taxonomy of dsDNA Phage Genomes at the Genus and Species Level. PHAGE. https://doi.org/10.1089/phage.2024.0050

Phage Therapy Applications · Phage Production and Manufacturing · Tools

We’ve all heard about Baylor College of Medicine’s TAILOR Labs for years (I’m convinced they’re almost single-handedly manufacturing compassionate phage preparations for the entire US right now…).

I still remember Austen Terwilliger, TAILOR’s Cofounder and Director of Operations, sitting across from me at a Phage Futures conference dinner in ~2019 saying ‘we’re going to do personalized phage therapy as a service. In Texas.’

My eyes got big — I assumed he didn’t know you couldn’t do that. He was very polite, but said ‘Well, that’s what we’re going to do’ and seemed confused why I was confused. (In the end, he was right!)

6 years later, TAILOR has treated tons of people with phages, is a mainstay in the phage field, and is depended on by countless patients and their doctors. And now, the group has spun off a company to scale up their strategy to more patients. That company is Phiogen Pharmaceuticals.

I was so excited to sit down with Amanda Burkardt, Phiogen’s CEO, to talk about what they’re doing, how they’re doing it, and most importantly…. in this economy??

Biotech is hard across the board, especially these days, but phage therapy companies have struggled from the added issue that infectious disease drug development isn’t really being done by pharma anymore… I’ve been reading about Achaogen’s bankruptcy (the antibiotic company that made a new FDA approved antibiotic, then went out of business right away anyway). Heartbreaking. How is a phage therapy company supposed to avoid that fate?

So this was why I reached out to Amanda. How does she think about the business case for phage therapy? How is she making the math work for her investors, when so many have decided against this sector? Do we need to price phages differently than antibiotics for it to be attractive enough of an investment? Otherwise, how are we ever going to bring phage therapy to patients? How are we going to treat their multi-drug resistant infections at all?

Luckily Amanda was game for this conversation, and she’d been thinking about it a ton too. And she brought along her COO, Dr. Mayukh Das, who actually worked on the iconic 2016 Patterson case, then on the wine grape phage cocktail developed by Texas A&M and A&P Inphatec, then at Johnson & Johnson as Associate Scientific Director of Phage Microbiology (back when J&J had such a department). So he’s been in the phage field for years, has been on the inside of large pharma, and he also took a bet alongside Amanda that they could make a phage therapy company work.

Joe and I were so excited to have them on the podcast, and they did not disappoint!

Highlights from our conversation:

• First — on how Phiogen differentiates itself. One way is it’s developed a "Directed Evolution Machine" that can enhance specific phage properties like host range, biofilm penetration, and immune stimulation.
• Is there a precedent for pricing antimicrobials higher than we typically see? Yes. Recent approvals of Live Biologic Products for C. difficile infections provide a pathway for phage pricing up to ~$10,000 per treatment. Will this be enough? We shall see, but it’s a start.
• US Centers for Medicare & Medicaid (CMS) wants to engage with phage companies much earlier than anyone realized; they recognize precision antimicrobials is a need, and that reimbursement discussions need to be had, now.
• The UK's new antimicrobial subscription model can provide guaranteed revenue (~$10M a year for 10 years) for certain antimicrobials. This has reassured investors Amanda talks to, and gives her some real numbers to back up discussions of Phiogen’s future economic viability.

A few snippets from the episode…

On investor skepticism:

Amanda: "We are paying for some of the sins of others that have come before us, but [also, this] has just not been proven. And so there's a lot of things that we're fighting against that are yet to be seen."

On technology differentiation:

Amanda: "We found a phage that was hyperlytic. It killed very quickly. But the host range wasn't as broad as we wanted it to be. And so we ran it through our [directed evolution] machine. And after four hours, we were able to evolve an offspring phage which had increased its host range by 10 to 15 percent."

On reimbursement pathways:

Mayukh: "The CMS specifically mentioned that they would prefer to be engaged in the drug development process as early as possible. So in that way, they can actually analyze the need of the specific therapy based on many factors like unmet need, your target population... and come up with some specific reimbursement model."

What's next for Phiogen:

Watch for updates on Phiogen’s recurrent UTI clinical trial planned for this year, and their collaboration with the patient advocacy group Live UTI Free to develop patient-centric phage products.

Listen to the full episode on the Podovirus Podcast!

• New: Listen on Apple Podcasts!
• Listen on Spotify
• Watch on YouTube

This episode is part of our series exploring phage therapy from multiple angles, especially trying to figure out what’s holding it back. Thanks to Joe Campbell, my co-host, for helping navigate these complex questions!

Please subscribe to Podovirus on Apple Podcasts or Spotify to support/encourage us, and to keep up with future episodes as soon as they drop!

Would love your thoughts!

• Add a comment to my tweet thread here!
• Add a comment to my LinkedIn post here!

Learn more

• John Rex’s financial post-mortem on antibiotic company Achaogen
• UK subscription model for antimicrobials
• Pasteur Act information (the US’s proposed version of an antimicrobial subscription model)
• Rubrum Advising is the healthcare advising company that organized the phage therapy reimbursement workshop that enabled Phiogen to talk to former leadership from Centers for Medicare & Medicaid about its views on phage therapy reimbursement — reach out to Neha Prasad at Rubrum if you’re interested in future workshops like this!
• A blog post Amanda wrote last year for Capsid & Tail on Phiogen's approach to phage therapy: "The Missing Middle Ground of Phage Therapy: Effective Phage Therapeutics that Bridge Economy, Regulation, and Unprecedented Functionality"
• The phage therapy center that led to Phiogen's beginning: Tailored Antibacterials and Innovative Laboratories for phage (Φ) Research (Baylor College of Medicine’s TAILOR Labs)

Intellectual Property and Commercialization · Phage Therapy Access and Implementation

Hi phage phans,

Look at that, we’re on Issue 300! That’s a really big number!

When we first started Capsid & Tail, we only planned to write maybe 30 issues. Years later, we’re still going strong!

Jessica’s still heads down working on Pf and ΦX174 phages at Stanford, while I’m heads down working on Evergreen and prototyping some other projects — there’s so many fun experiments to share, but still so many bugs to fix.

I’ve also been messing with the Phage Directory website, and we’re ready to flip the switch to the new site soon. Excited to share!

Here’s what we published this past month:

First food, then pharma: How Intralytix played the long game with phages

by Jessica Sacher

Jessica recaps her podcast conversation with Sandro Sulakvelidze about Intralytix’s 25-year-long journey, from startup to one of the few profitable phage companies.

Sandro shares how growing up in phage-friendly Tbilisi led him to found Intralytix after witnessing US patients dying from infections that could have been treated with phages, and explains their strategic pivot to food safety applications before moving into human therapeutics. The conversation covers Intralytix's experience with FDA approvals, their AI-driven manufacturing innovations, and Sandro's vision for the future of phage therapy.

Intralytix: https://intralytix.com/

Evergreen Phage Meeting hits the road for 2025

by Ria Kaelin and Tom Denes

Ria and Tom announce that the Evergreen Phage Meeting now in Knoxville, Tennessee for 2025! After 50+ years in Olympia, Washington, the meeting is expanding its reach while honoring Dr. Betty Kutter's legacy and introducing new scientific tracks including diagnostics, regulation & policy, biomanufacturing, and career development alongside traditional phage biology topics.

Sign up before March 2nd for Early Bird! https://evergreen.phage.directory

Building a phage lab: a journey of resilience, teamwork, and sticky situations

by Silvia Würstle, Maria J.G.T. Vehreschild, Simone Lieberknecht-Jouy, Fedja Farowski

In this article, Dr. Silvia Würstle and her team share their experience transforming an abandoned, water-damaged space into a modern phage laboratory at the University Hospital in Frankfurt, Germany. The team describes their journey from dealing with warped floors and empty rooms to creating a fully-equipped, certified lab space, highlighting challenges like navigating hospital ordering systems without product photos, securing specialized equipment, and learning technical specifications on the fly. Despite the obstacles, they successfully built three lab spaces, an office, a common room, and a storage area with support from hospital staff and the global phage community.

Silvia’s new lab website

Setting up systems to make phages available for all

Throwback: I really loved this article from last year, where Jessica laid out our entire process for setting up our phage lab’s system. I think this article pairs very well with both Sandro’s and Silvia’s Hero’s Journey stories!

~ Jan & Jessica

Intellectual Property and Commercialization · Phage Community and Collaboration · Phage Production and Manufacturing

Setting up a new laboratory is always a challenge. But setting up a bacteriophage lab in a previously water-damaged space that had been unused for ten years? That’s a story for the ages - and one filled with equal parts humor, hard work, and heartfelt gratitude.

Our journey began in Frankfurt, Germany, within the Department of Infectious Diseases of Prof. Maria Vehreschild at the University Hospital. The rooms earmarked for the new lab hadn’t seen action since a catastrophic water incident left the floors looking like rolling hills. A complete renovation was in order, including ripping up the old flooring and starting from scratch. What awaited us after the dust settled? A blank canvas of tables, sinks, and dreams - without a single piece of lab equipment.

A shopping odyssey

Equipping the lab became a full-time sport. We gathered an almost comical number of quotes from suppliers and spent countless hours sweet-talking companies to nudge us higher up on waitlists. Through persistence (and daily phone calls), we secured some fantastic devices and robots that have truly transformed and streamlined our lab work. One of our victories: a shiny new TapeStation, ending the era of electrophoresis gels that had a bad habit of overrunning while Silvia was stuck in yet another long meeting. In addition, for >500 material items of the lab, we had to comb through the hospital’s internal ordering system, which features no product photos - just one short headline without description. From filter tips to magnetic stands, it was a game of memory and deduction.

Fig. 1. Demonstrating examples of what not to do in the lab.

Knowledge, handymen, and microsiemens

The organizational skills we developed in our previous labs, from managing orders to handling repairs, proved invaluable for this venture. That experience was essential in identifying the devices that would best suit our phage research. Of course, not everything had been learned in advance. Case in point: when a technician asked Silvia what conductivity level in microsiemens/cm she preferred for the water purification system. Her perplexed expression - oh, how we wish we had a camera ready for that moment!

We tackled everything from installing new gas lines for cell culture to selecting the exact right wall paint to meet genetic engineering certification requirements. Thanks to incredible support from the whole hospital and regulatory authorities, we achieved further certifications for BioSafety Level II pathogens and animal disease agents. Along the way, there were painful lessons - like discovering the hard way that the sticky plate in our shaking incubator wasn’t quite sticky enough for certain racks with bacterial cultures.

Fig. 2. No more sticky plates in our shaking incubators.

The people who made it possible

The result? Three sparkling lab spaces, one bustling office, a cozy common room, and a well-stocked storage room. Most importantly, we now have an unbelievably brilliant team that brings this space to life every day. We can’t imagine a better group to work with.

Fig. 3. The new lab!

This entire venture would have been impossible without the collective effort of so many people. From the tireless craftsmen and administrative staff at the University Hospital to the global network of phage scientists who shared their advice - thank you.

Next time, maybe we’ll hire a lab planner. But for now, knowing every inch of this lab and every beep of the machines is proving to be a secret superpower. Here’s to many discoveries ahead!

Learn more about the Wurstle lab

• Lab website

Personal Experiences and Reflections · Lab notes

Evergreen has never been just a conference — it is a catalyst for innovation, a gathering place for pioneers, and a driving force behind the future of phage science. For over 50 years, it has brought together researchers across disciplines, creating a home for discovery, mentorship, and real-world impact.

In 2019, Evergreen was the largest it had ever been. In 2021, we refused to let a global crisis halt progress, launching our first hybrid meeting to keep research moving forward. In 2023, we said farewell — not to Evergreen itself, but to the location that nurtured us for so long.

Now, we look ahead — not to an ending, but to an era of expansion and reinvention.

Knoxville: The First Stop on Our Journey

Evergreen is hitting the road. As part of our new rotating format, we’re taking the meeting beyond Olympia, allowing us to expand our reach and highlight leading institutions in phage research worldwide.

Knoxville, Tennessee, is the perfect place to launch this new era. Dr. Tom Denes and Dan Bryan, both former students of Dr. Betty Kutter, now call the University of Tennessee, Knoxville (UTK) home, where they continue their work in phage research. UTK faculty are eager to join us in shaping the future of Evergreen, providing an ideal setting for this next chapter.

When the Phagebiotics Research Foundation decided to take Evergreen on the road, we knew we needed a strong first host to support us as we experiment, grow, and ensure a smooth transition into this new phase. Knoxville gives us that foundation, and we are excited to begin this journey with them.

Expanding Our Scientific Focus

Phage science is advancing at an unprecedented pace, yet challenges remain. Regulatory landscapes are shifting, access to phage therapies is still limited, and promising solutions have been overlooked—only to be rediscovered and validated in modern times.

Evergreen has always been a place to tackle these challenges head-on, and 2025 will be no different. We are expanding our scientific focus with new tracks that reflect the rapidly evolving landscape of phage research:

Phage Biology, Evolution, Ecology, and Therapeutics – Continuing the core tracks that built Evergreen’s legacy.

Diagnostics – Phage-based tools for infection detection and microbiome analysis.

Regulation & Policy – Navigating the evolving landscape of phage therapy approvals and global standards.

Biomanufacturing & Scale-Up – Bridging research and industry to bring phage-based products to market.

From Classroom to Career – Connecting students, postdocs, and industry leaders to foster professional growth.

This meeting has always been about collaboration, mentorship, and advancing the field. As we step into this next phase, we invite you to participate. Bring your research, share your ideas, and help shape the future of phage science.

Honoring Dr. Betty Kutter

Evergreen would not exist without Dr. Betty Kutter, who dedicated her life to mentoring young scientists and championing phage research, even when the field was at the margins of mainstream science. She made sure this community had a place to grow.

Now, as Emeritus Leader, Betty is enjoying her well-earned retirement in the Pacific Northwest. Though she won’t be with us in person, her legacy is in every researcher, every discovery, and every connection formed at this meeting.

Evergreen 2025 is a tribute to her vision and dedication; we are honored to carry it forward.

Join Us in Knoxville

Evergreen is more than a conference—it is a movement. It is where ideas take shape, collaborations begin, and the next generation of scientists finds inspiration and opportunity.

We hope you’ll join us, share your work, engage in discussions, and contribute to the future of phage research.

Register now and submit your abstract, at:

https://evergreen.phage.directory

We look forward to welcoming you to Knoxville for Evergreen 2025: Where the Future of Phage Science Takes Shape.

With deep appreciation and anticipation,

Tom Denes & Ria Kaelin Co-Chair Organizers, Evergreen Phage Meeting Phagebiotics Research Foundation

Phage Community and Collaboration

From a chance conversation about a patient lost to antibiotic resistance, to pioneering FDA-approved phage products, Dr. Alexander "Sandro" Sulakvelidze shares Intralytix's 25-year journey in our latest Podovirus podcast episode.

Growing up in Tbilisi, Georgia, where phage therapy was common, Sandro was shocked to see patients in the US dying from infections that could have been treated with phages. This realization led him to found Intralytix in 1998.

Intralytix's unconventional path — starting with food safety applications before moving into human therapeutics — offers several lessons for the phage field. Today, Intralytix stands out as one of the few profitable phage companies (if not the only one?), with multiple FDA approvals in hand for food, and now some key ongoing human clinical trials.

Highlights from our conversation:

• Sandro has watched the phage field has transformed from complete skepticism to active engagement, with hundreds now attending phage conferences
• Intralytix wasn’t always food-focused; Sandro shares their strategic pivot to food safety applications and his thoughts on how this was crucial for their long-term survival
• How Intralytix worked with the FDA to get approvals for its phages, and Sandro’s views on regulatory barriers that exist for phages now
• Sandro’s views on how AI and machine learning are revolutionizing phage manufacturing, and how the system they set up now has 100% (!) success in predicting optimal production conditions for their phages
• How current regulatory concerns about transducing phages may be disproportionate given natural background levels (Sandro does math on the air!)
• Infrastructure for clinical phage therapy remains a major challenge, but why Sandro sees a ton of hope

A few snippets from the episode…

On the evolution of regulatory approval:

Sandro: "My experience with FDA has been very positive. Both CFSAN and CBER have been reasonable to work with. We were the first company to obtain FDA approval for phages in food safety applications — that was a historic moment that finally broke that barrier. We now have five FDA-approved products on the market for food safety."

On manufacturing innovation:

"The variables you have to examine are mind boggling. What's the best MOI? What's the best harvest time? What are the aeration requirements? Using AI funded by NIAID, we've summarized 25 years of production data to predict optimal conditions. So far, we have 100% success — whatever the program predicts, we achieve in real life."

On infrastructure needs:

"There is no infrastructure for phages currently in the medical community. Every major hospital has the ability to quickly determine bacterial susceptibility to antibiotics. There is no such thing for phages. You need instruments, trained personnel, and standardized testing methods. This will be one of the biggest challenges."

Looking ahead…

Sandro’s 5-Year Vision:

• Continued focus on personalized medicine approaches
• Possible emergence of 1-2 mainstream products, likely through accelerated pathways
• Growth in nutraceutical phage products
• Gradual infrastructure development for phage therapy

Sandro’s 15-Year Vision:

• Multiple mainstream products on market (contingent on efficacy proof)
• Enhanced infrastructure including phage reference centers
• Expanded use of phages in microbiome modulation for conditions beyond infection
• Established reimbursement pathways

Jess’ reflection on the conversation

This was a great conversation, and I so appreciate Sandro’s openness about this entire quarter-century adventure he’s been on! This quote from Sandro stood out to me:

"Deep in my heart, there is no doubt this works. There are just too many successful stories."

And yet, he acknowledges that this is not enough; the field's future hinges on demonstrating efficacy in rigorous clinical trials.

I think this sums up the stance I hear most from phage professionals everywhere… we have a deep sense that we wouldn’t have so many successful case studies if phages didn’t work… but we know that’s not the evidence we need. What we need is a successful phase 2 clinical trial, stat! So thanks to all of you out there who are persevering toward this goal.

P.S. If you think your team is close (or you’re about to give up), come on the podcast and let’s chat!

What's next for Intralytix:

Watch for results from Intralytix's Shigella trial by mid-2025. This NIAID-funded study uses an FDA-approved challenge model and could provide crucial efficacy data the field needs.

Further reading:

• Watch the full episode on Youtube
• Listen on Spotify (and subscribe to get the next ones!)
• Learn more about Intralytix's food safety and therapeutic products

This episode is part of our series exploring phage therapy from multiple angles. Thanks a million to Joe Campbell, my esteemed co-host, for helping steer this conversation to interesting new places, as always!

Stay tuned for an upcoming episode on drug pricing and regulatory pathways, with Amanda Burkardt, CEO of Phiogen Pharma!

Intellectual Property and Commercialization · Products · Phage Production and Manufacturing · Regulatory and Ethical Considerations

Hi everyone,

Time flies, doesn’t it? I still remember writing the outlook letter in 2024. We were still in Australia, still thinking about our big move to SF. Researching how to move our adopted kitty Mivi across the pond. A year later we feel like we’ve lived in this city for decades. Oh how time flies!

For 2025, Phage Directory (and Capsid & Tail) will largely be the same. Capsid will still be covering phage news, jobs, and events (please send your jobs and conferences to us!).

For this year, I’m eager to cover more papers in the intersection of machine learning, biology, and phage, as both Jessica and I believe that phage therapy research will continue to drift towards becoming more data-driven and computational. While I will cover more ML/bio papers, Jessica is planning on releasing many more Podovirus podcast episodes.

(If you’d like to be a guest on the podcast, email us!)

As for guest posts, we’ll have an upcoming post by:

• Silvia Würstle, on how she got her new lab set up as a PI, her tips and tricks, and what she learned;
• A podcast interview with Ross Rheingans-Yoo, a biotech investor who thinks phages are actually very interesting, and who’s on a mission to make clinical trials cheaper;
• A recap of my interview with Sandro Sulakvelidze from Intralytix, on how he spent the last 26 years building a long-lasting phage company, and how Intralytix is about to get results on its phage therapy challenge trials (heathy human patients infected on purpose with Shigella, then treated with phages, here in the US!)

One of the biggest goals this year is to add more resources — like tools, guides, databases, etc. — to help phage labs get more computational. I’ve been playing around with adding more resources like bioinformatics and ML, but I’ll make an attempt to double down on it this year. We have some ideas, but we’d love to hear from you.

As part of this shift, we’re taking Phage Directory back to its roots — as a scientific resource and community for phage therapy researchers. I’m also hoping to add some tools to make it easier to search for phages, bacteria, and phage papers. Stay tuned!

The biggest delight of 2025 is how the phage conference circuit is absolutely tingling with options, from the International Symposium on Phage Therapy in Portugal, hosted by Joana Azeredo and her team, to the Biennial Phage/Virus Assembly Meeting in Cape Cod, to the IAS-USA Conference on Bacteriophages in DC, run by Graham Hatfull, Chip Schooley, Paul Bollyky, and others.

And of course… drum roll… the return of the Evergreen Phage Meeting! This time in Knoxville, Tennessee.

Evergreen’s website isn’t ready for signups yet (I’m still squashing bugs and we’re finalizing some details, but I see all of you who figured out the link and are signing up anyway. You’re the true fans. I see you!)

Again, we’re super excited for 2025. There’s so many conferences, and so little time.

My prediction for 2025: by the end of 2025, at least 10 phage labs (that have never previously touched machine learning) will build and train a useful model based on their data.

Jessica’s prediction for 2025: We’ll see an array of phage therapy centers popping up, especially across Asia (China, Singapore, Nepal, etc) but also in Europe and potentially South America (Uruguay is already almost there!). Clinical trials on phage therapy will start springing up especially in China, and they’ll move fast. We’ll see multiple started and completed phase 2 phage clinical trials by end of 2025. At least one will show efficacy!

To phages and beyond!

~ Jan & Jessica <>-{

Phage Community and Collaboration · TechBio

"Hi everyone,

Thanks for making it this far — welcome to the last Capsid of the year!

I don’t know about you all, but this year has hit a bit differently for us, for better or worse. We’ve been trying new things, been trying to find our footing, been trying to settle in, but instead of settling in, the year has remained quite un-settling.

There’s a lot we looked forward to this year, from the city, from ourselves, from the unknown, and I think looked a bit too forward and fell off the deep end. Maybe our expectations were a bit too high?

We’ve been doing — trying — a lot of things. We’ve definitely learned a ton. If we’d merely expected us to learn new things then by all means, we’ve blown past that metric. Paradoxically, I think we tried more things this year than we’ve previously tried, but at the same time, I feel like we haven’t gotten as much done as we’ve hoped. There’s a lot of experiments we tried but scrapped along the way. But maybe that’s what comes with the territory. If your experiments all go right, you’re not experimenting hard enough.

With that said, there are still plenty of things we’ve done this year, that we’re proud of!

~ Jan

Our favorite Capsids

Ok to start off, one of my favorite Capsids posts is by far the March Phage Picks. This is kind of a cheat, because the post itself isn’t actually my favorite Phage Picks, and those papers aren’t actually my absolute favorite papers. I just liked that we tried something fresh — and according to our email analytics (very few have actually written to use about liking this new format…), readers seemed to like it!

Our Phage Picks for March 2024!

This was a format I’ve wanted to do for quite some time. Both Jess and I have a huge log of favorite papers, and we’ve always wondered how to share our reading stack. About once a month, we share our favorite papers. Mine skew bioinformatics and machine learning, while Jess skews more phage biology. We have lots of ideas of how to improve on the format next year (plus figuring out ways to encourage our readers to share more of their favorite papers, as that’s been the hardest part…).

Building a SEED library for One Health

My other favorite article was this one from Jonathan Moore. It’s just so out there and not like anything else I’ve read in phages, but if you read the article a lot of it is grounded in lots of our existing understanding of biology, and it’s not actually that science fiction. I like this article because it reminds us that there’s a whole lot more applications out there for phages other than therapy.

~ Jan

Phage therapy is unfettered by borders

I loved this article and what it represents. This is an account from a Hungarian PhD student who helped screen and find phages for a patient we were trying to treat in Australia before we left.

~ Jess

PhageBase: A visual atlas of phages

This was a super cool initiative from an undergrad student I worked with in the Bollyky lab this year; he created this all on his own, and has a small team of fellow students manually curating TEM images from phage papers.

~ Jess

Our favorite Papers

I’ve been spending a lot of time reading up on and trying to understand machine learning this year. I’ve also been trying to read up on the space between machine learning and phage biology of which there are a surprisingly high number of both interesting and accessible papers.

In Phage Genome Annotation: Where to Begin and End, Anastasiya and Andy Millard writes the best paper to figure out where phage bioinformatics is today, and where it’s headed.

If you’re new (or catching up) with phage bioinformatics, you absolutely need to start here. Few papers blow my mind anymore, but this paper Deep model predictive control of gene expression in thousands of single cells changed how I thought about ML’s relationship with biology.

And if you’re wanting to figure out phages and machine learning, you absolutely need to read Dimi’s paper Prediction of Klebsiella phage-host specificity at the strain level, where it’s not just a paper but basically an instruction manual on how to set it up. It even includes very well-documented code, and all the data, for you to replicate over a weekend. He makes every piece so incredibly easy to understand.

~ Jan

I too have been trying to understand machine learning this year. Hard not to when you’re living in the physical global hub of AI, and every other week there is a meetup where you meet people working on applying it to biology. And when each day you’re using better and better AI tools for everything else… it’s hard not to start thinking it could be mega powerful if applied to actual scientific problems. (FYI if you’re not using Claude for reading/summarizing/thinking/writing/editing in science, and still relying on ChatGPT, you are missing out).

I personally got really inspired by the Evo paper by Eric Nguyen and colleagues at the Arc Institute (next door to us at Stanford!). The paper used phage data, without labels (ie annotations), to train a model that could predict new made-up protein-RNA complexes, like CRISPR and transposons, that actually work in the lab, and to even predict new phage genomes (that don’t exist, but should be functional, based on what it learned about the other genomes). That lab is now trying to synthesize these genomes to see if their model really does predict the rules for a phage genome. They are starting with phiX174, which is a small and simple one (the first phage/organism sequenced, I believe?).

This led me to Dimi’s paper (the one Jan mentioned above) on predicting strain-level phage-host interactions for Klebsiella, and then the one by Aude Bernheim’s group, led by Baptiste Gaborieau, doing something similar for E. coli. I already wrote about these in last week’s Phage Picks, but needless to say, they inspired me to start really digging into all this stuff. I even made phage + AI (and reorienting our research around using these new tools) into a focus of my recent talk at Phage Option in Colombia (which was an AMAZING time by the way; think the Evergreen of South America…. such lovely, lovely people and a relaxed, sunny vibe!). Anyway, much more on that soon; for now I need to spend Christmas break watching Coursera/Kaggle courses on machine learning 101 so I can figure out HOW to make it useful for phage projects. Stay tuned… and let me know if you’d be interested in joining a ‘Phage ML’ club where we help each other learn this stuff and start designing experiments/projects that generate data that can be fed into ML models. I really think this is what’s next for the phage field.

~ Jess

Jess’s year in reflection

This year I worked with Paul Bollyky’s lab to try to get a center for phage therapy and phage delivery off the ground (both getting it going physically, and getting the funding for it). His lab is interested in using phages for patients but also for delivering therapeutic genes to eukaryotic cells. I was super excited to join, because I wanted to a) see if I could set up a phage therapy center a second time, to prove the model of ‘academic lab + a set of protocols = patients treated’ is repeatable, b) learn how to write NIH grants from someone with a history of getting them, and c) work with really smart people who would push me to think bigger and take my work to the next level.

Walking up to the iconic Stanford campus…

Well, it’s been a pretty crazy year. I haven’t accomplished everything I wanted, but I’ve learned a lot, as Jan mentioned above. Here’s a few thoughts on the year:

Lessons in unseen advantages and unwritten requirements

I definitely learned a lot this year about the unwritten requirements for phage production for patients. Coming into this year I thought it was mostly about knowing what to do, and having the freedom to do it. I felt confident I had both of those things coming into this new role at Stanford — and I do! But I’ve found there are many additional ingredients to a ‘phage therapy center’ (my working definition: a lab that repeatably makes phages for patients) that I didn’t account for. Things I happened to have access to in Australia that has taken me longer than I anticipated to find/establish here (and would be the case anywhere, I think). I’m working on making this all a reality here, but here’s a few things I think should be added to the list of ‘phage therapy center’ ingredients:

Full time personnel dedicated to phage therapy lab work: In Australia I wasn’t writing grants or papers, wasn’t mentoring students. I was full time dedicated to phage therapy production. I was also teamed up with Stephanie Lynch, a phage PhD dedicated to the lab-side of setting up our phage therapy system, who was willing to spend time around the clock in the lab. Not only did she create at least half of the systems, she was there to bounce ideas off regarding labeling, banking phages, workflows, etc. This gave me so much time (and energy) to think about what to do, document what we were doing, and write about it, and the work didn’t grind to a halt if I wasn’t personally able to be in the lab much on a given day (or if my 1.5-hour train commute turned into a 3-hour one a few times every couple weeks due to spontaneous delays). In short, a phage therapy center needs a Stephanie.

Lab space & equipment dedicated to phage therapy work: Lab space is always hard to come by, and that is especially true here (you are actually not allowed to build new buildings on this campus without knocking one down, I have heard?). And there are many, many enthusiastic students and postdocs to share each square meter with. Again, probably the norm in most academic labs. In Australia we were spoiled: we had multiple rows of dedicated bench space for phage production, new state of the art purification equipment no one else was using. For phage purification, for example, I didn’t have to contend with much shared equipment, which meant when I left a machine, I knew no one would use it between my runs. I didn’t have to implement as many checks for cleanliness (ie. is there phage circulating in the AKTA from the last person’s run? Have to check before every run, which means a titre + extra waiting day before any new run). I also had a validated, calibrated endotoxin reader available that I could use, so I didn’t have to validate biochemical assays myself to be able to trust an endotoxin reading. Anyway to sum it up, having a bunch of empty space and pristine new machines no one else needs to use, plus a free human therapeutic quality control lab with professionals who maintain the equipment, is not really a reality in most academic labs…

Quality control personnel/facilities: As mentioned above, in Australia we had walking-distance access to a CAR-T therapy center to do our quality testing (and they were fine with us bringing phages in there!). We had a hospital micro lab down the hall that was happy to do our sterility testing. And we had a bioinformatician and sequencing team dedicated to helping us set up a QC pipeline for patient-ready phages. I’m sure these types of resources do exist here at Stanford (I’m new, so of course it takes time to identify and vet these resources). But so far I haven’t been able to convince the CAR-T centers here use their equipment for our phages, and haven’t yet attempted to see if the hospital lab will do our sterility testing. The students in this lab who did bioinformatics graduated just as I arrived, and my beginner Geneious & coding skills are hardly sufficient to make much progress fast. Sequencing facilities of course exist here, and we have a lab upstairs willing to do it for us, but turnaround time has been challenging, and setting up a pipeline for this that you know works for your purposes (e.g. how do we know we have all the reads we need to make a call about sample purity?) without dedicated bioinformatics support has been hard.

So this year for me has been a lesson in unseen advantages. Things about your current setup that you don’t think about, but are part of the reason for your success. When you have to fit in ‘set up a phage therapy pipeline’ between writing, mentoring students, and other research projects and meetings (no doubt the case for 99.9% of academic labs), phage therapy for patients becomes an uphill, almost unfathomable battle. It becomes hard to imagine how to do it safely. It makes me feel sheepish that I ever went around saying ‘an academic lab can do this’. We weren’t really a normal academic lab over at Westmead. It makes me even more grateful for getting to be a part of that over there, and makes me realize there’s a lot more unseen components to scope out and name before I can deliver a ‘phage therapy center kit’ on a platter and expect a normal academic lab to be able to run with it. I’m so grateful that this current experience let me learn that!

Of course, I’ve only been in this lab for 7 months. There’s a ton more I could still do to get a phage production system working. I certainly have the freedom to do it; a supportive PI and labmates. Stanford has a ton of resources, no doubt; I just probably need more time to figure out what and where they are, more time to meet the people who can help me surmount some of these hurdles, etc. And just more time to pause and articulate the needs and bottlenecks (this post is helping me finally do this!) so I can unlock what I need to solve them.

All in all, I have still made strides; I’ve started getting a phage biobank set up, establishing a sequencing pipeline, getting our phages sequenced, implementing a database of our phage stocks and batches, a batch labeling system, logbooks for our repetitive experimental work like titres and propagation runs. I’ve learned to use our AKTA Purifier and OH-chromatography (which works great!), figured out what it takes to clean it (and prove it’s clean) so it can be reused for new phages each time, and I’ve evaluated three endotoxin testing methods (to save you all the time, the best is what we were using in Australia; every phage lab should just buy an Endosafe® nexgen-PTS™ endotoxin reader and be done with it). Lastly, I’ve gotten to work with some truly amazing students here; it has been absolutely mind boggling to see how self-motivated, fearless and fast-learning Stanford students are! (They write and submit their own research papers and genbank sequences without help, learn how to use new equipment on weekends, and constantly make me question if we ARE going too slow with our progress…)

Overall, the experience of trying to do the same thing in two consecutive settings, each with their different strengths and limitations, has given me a ton of new perspective I didn’t have before. I know this will serve me going forward as I try to support others getting phage therapy off the ground in their institutions. This year I’ve felt at times like I’ve been spending 100% of my effort to get 10% of the way toward something I already achieved once, and this has made me question if I’ve been going about this the right way. Maybe instead of trying to copy what works in one setting, we should look at what we each do really well, and have unique access to, and build on that. Stanford has such cutting edge research going on, so much proximity to AI and machine learning experts, access to research funding, and access to incredible students. What could we do here, that we couldn’t do elsewhere? This is where my head is at lately. To be continued on where that leads!

Lessons in grant-writing

One goal I’m proud to have accomplished this year was to learn about grant writing — I got a TON of hands on experience with several grants, and feel like I can understand the anatomy of how they work in this country (at least for the NIH). I worked on 3 center-level (ie. ginormous multi-PI, multi-project, sometimes 600-page) NIH grants and helped respond to reviewer comments/resubmit a handful of R01 grants, which was enlightening too. I learned there’s a whole suite of rules about what makes a good one — more on these learnings in a future post! (And places like Stanford actually teach their students and staff courses on how to read between the lines and structure everything exactly right — not to mention giving labs full-time grant writing staff to help them and take care of the admin aspects… talk about unseen advantages!). Very grateful for the mentorship I’ve received from Paul on the grantwriting process, and for how much I was able to get exposed to in a short time. I also got to feel the pain of not getting two of them (so far). Ouch. It’s not fun, but it’s interesting to see the comments and the feedback you get — it’s way more detailed than I would have thought. I could see how you could build on this and how it would become fun...

Lessons in review writing

I am also proud of a review I submitted on phage therapy this year (it just got accepted today, yay!). My first ‘last-author’ paper, which feels pretty neat. Shoutout to the huge team of students and postdocs who each contributed parts of this (Saumel Perez, Grace Cullen, Zhiwei Li, Kevin Chen, Tejas Dharmaraj, and Tony Chang), to the collaborators who helped shape it (Gina Suh, Rob Lavigne, Sabrina Green, Jean-Paul Pirnay), to Paul Bollyky for instigating it, providing mentorship on how to manage the whole thing, and keeping us all motivated, and to Kevin Kim for driving it forward and doing the vast bulk of the work. And to Arya Khosravi for the support behind the scenes; offering help and edits and ideas.

Bollyky lab labmates! Some of the best people I have met. This was Halloween when we dressed up as minions… I still have not seen Minions or adjacent movies BUT turns out that wasn’t a requirement.

Lessons in podcasting

Lastly, I am SUPER proud that I answered my own nagging curiosity and got a podcast off the ground this year: Podovirus. I was both curious about the medium (would I like it as much as I like listening to podcasts? So far yes) and about the content (WHY don’t we have phage therapy yet? What is the hold up? This is driving the conversations and choice of guests so far). We’ve just released the 5th episode, with Sandro Sulakvelidze of Intralytix, and have 3 more that are recorded and mostly edited; almost ready to go. This has been in collaboration with Joe Campbell, who retired from NIAID this year, and he’s been fantastic to work with! It’s been a ton of work but I’ve loved exploring this new medium. More to come in the new year!

That’s all from me for now! I hope everyone has a wonderful holiday season, and I hope you each get and take a real break. I will be in Calgary with Jan and my family, relishing the fact that they are 2.5 hours away instead of 16, which was the real reason we moved our lives back across the pond after all.

~ Jessica

Jan’s year in reflection

So while Jess have been getting back into lab and grant work, what have I been up to? Well, I was going to upgrade Phage Directory but instead I got distracted. Now to be fair, this is basically every year, that’s why Phage Directory’s code is kind of in shambles, and lots of things are broken, and I make the promise that next year is the year I fix and rewrite all of Phage Directory.

My main distraction this year was everything AI. We even launched a new effort separate from Phage Directory to explore these ideas further, at labspace.ai, where we build experimental data and interface tools for data management, bioinformatics, and automatic Phage Directory and Capsid & Tail.

labspace.ai is a collection of experimental data management and research tools. Take a look — they’re all free to use!

Up until this year, a lot of stuff in C&T was manual — from cropping and resizing cover images to copying and pasting text into Mailchimp. This year, we built a lot of data flows to automate that for us. As you probably noticed, we also built a lot of tools that helps us tweet, summarize, and help us understand articles with science communication. Behind the scenes of Phage Directory we’ve changed almost everything; swapped out the API, upgraded the database, etc. which allows us to update the way we’re handling profiles, integrate next year’s Evergreen directly into Phage Directory, and a whole lot more. On the side of doing phage data, we’ve taken what we’ve learned from Phage Australia and built a bunch of tools for biobank data collection, which we haven’t rolled out yet, but includes neat features like a phage/host range heat map visualizer and Circos-like chord diagram system, to draw phage-host relationships (or co-authorship/collaboration relationships!). I’m even excited to add a better search mechanism, add better filters on Capsid, group special issues like Phage Picks, and be able to just read articles without the What’s New parts. Oh, and I’d absolutely love to build tools no one’s asked for, like a way to chat with Capsid & Tail 😉.

Since a lot of the upgrades have been on the side of code, it’s been really hard to show what’s going on — and if I wrote a post about it without the tools being fully available, they’re just kind of fruitless demos. Most of the systems are now in-place though, so next year we’ll definitely be making some sweeping changes around Capsid & Tail and Phage Directory.

Instead of announcing and promising a bunch of things next year, I’ll just be building those out over the next month or so, and hopefully you’ll just start noticing those changes throughout Phage Directory!

One thing we’re looking forward to next year is to just simplify-simplify-simplify and go back to the roots of Phage Directory — back to being a cozy close-knit community of phage researchers. And Evergreen. We’re definitely doing Evergreen again next year!

~ Jan & Jessica"

Personal Experiences and Reflections · Phage Production and Manufacturing · Tools

Hello everyone!

It’s been a while since I’ve done one of these (have I ever?). It’s mostly been Jan’s thing. But lately I’m really digging into phage + ML (even though I am a complete ML newb). Some of you may have seen our talks in Cartagena, Colombia last week (hello new South American phage friends who’ve joined us here at Capsid & Tail this week!). Anyway if you saw them, you’ll not be surprised at my picks this week. I talked about them in my talk… namely, I think we should pay attention to what is going on with building new ML models with phage data. Hint: it is no longer just phage people. The labs near us at Stanford now are training generative DNA models on phage data, and doing crazy things like generating whole new genomes! What is next? Can we (should we) design all our experiments so that machines can gobble it up and learn more stuff faster? How would we even do this? I will be exploring this further in coming weeks/months. But first, let’s look at some of these papers!

I also chose a paper that’s about actually CHOOSING projects. It came up today in conversation with an undergrad student I am mentoring in the lab, Aaryan, who was wondering what he might do in January in the lab. My answer was, I could come up with something, OR you and I can each read this paper, and then meet up to discuss. TLDR is that we should all be spending more time thinking about what to do before we start lab work. We currently spend miniscule amounts of time picking a project, and then years answering to it. Why?

Hope you enjoy!

~ Jess

Sequence modeling and design from molecular to genome scale with Evo

What is it about?

Everyone’s using ChatGPT, but how long until we get ChatGPT for DNA? It was always inevitable, but who knew it would be so soon? This paper introduces Evo, a new generative machine learning model built by a group at the Arc Institute (a new institute next door to Stanford funded by Silicon Valley billionaires who want to give biology a boost — this is one of their first big releases and it’s awesome). A team there trained Evo using about 3 million prokaryotic and PHAGE genomic sequences (just the nucleotide sequences… not annotated or labeled!). It can predict and GENERATE functions across DNA, RNA and proteins. For example, it can generate CRISPR-Cas protein-RNA complexes and transposons that don’t exist (but function in the lab!) from scratch, and create plausible phage genome sequences over 1 megabase in length (though they haven’t yet rebooted these genomes, they told me they are close). Here is a seminar Brian Hie, the last author on the Evo paper, gave a little while ago; I watched this before reading the paper and it really made it click for me.

Why I'm excited about it:

I am obviously not an AI/ML expert, but even I can sense that this a huge leap forward in applying AI to biology. The fact that Evo can predict complexes of protein and RNA, AND do this all from nucleotide sequences alone (without even being labeled/annotated)… this seems wild to me. To me it means that a lot more of biological information may be encoded in the DNA than I thought. And if this was done with just nucleotides, without labeling what all the genes are (’this is a CRISPR sequence, that’s a transposon, that’s a chaperone’), then imagine what it could do once we (somehow?) add in all the knowledge we have amassed as humans about biology!

Secondly, being able to generate genomes would be pretty huge. At first I was somewhat indifferent when I heard this team was ‘trying to create the first synthetic phage’, because I thought it was already done, ie. synthesizing the DNA from a sequence, using something like Twist, IDT or Genscript. But no, this is generating a phage genome sequence that doesn’t exist, then checking if it would work if it were synthesized (by synthesizing the parts, sticking them together, then putting them into a cell to ‘reboot’ the phage). Does this mean I could finally say ‘I want a phage that goes to the bladder and kills all the UTI bugs once it gets there’, and even though I don’t KNOW the features the phage would need to have (non-immunogenic capsid proteins? some sort of not-yet-discovered tag that makes the kidneys let it pass it into the bladder? cross-genus host range?), I could generate a genome that would lead to a phage with these features?! (I asked Brian Hie and he said something like ‘not yet but essentially yes’).

Lastly, the model has been released openly, so researchers anywhere can build on this work and take it in new directions. In fact, a hackathon team (Team PhageBook!) did just that to build a phage-host prediction system using the Evo model as a base (in 10 days)! This team used Evo to do phage-host prediction that seems to rival some of the new papers coming out backed by a whole bunch of phage bioinformatics and expert annotation… (More on this in a future Phage Pick I think!). ~ Jess

Paper: https://www.science.org/doi/10.1126/science.ado9336

Nguyen, E., Poli, M., Durrant, M. G., Kang, B., Katrekar, D., Li, D. B., Bartie, L. J., Thomas, A. W., King, S. H., [...], & Hie, B. L. (2024). Sequence modeling and design from molecular to genome scale with Evo. Science, 386(6723).

Prediction of strain level phage–host interactions across the Escherichia genus using only genomic information

What is it about?

More phage machine learning! Now I am hooked, I have started to seek these papers out. This paper is by Aude Bernheim’s group in France, and focused on understanding and predicting how phages interact with E. coli strains. The team assembled a collection of 403 E. coli strains and 96 phages (I was surprised that it wasn’t THAT many; I was thinking there had to be like, thousands? tens of thousands?), systematically tested how they interact, and loaded up the data into machine learning models. And voila! They were able to predict interactions with 86% accuracy just from genetic data alone. They also developed an algorithm for recommending phage cocktails that could target specific E. coli strains, and their cocktails worked! One super interesting finding was that, based on which of their model experiments worked the best, they could find out that phage binding to bacteria matters more for successful infection than the bacteria's defense systems. This was also recently shown in this paper for 2 phages, done through GWAS/transposon mutagenesis methods. Interesting! How often / for how many phages and species is this true, I wonder? And what else about phage biology can building these models teach us?

Why I'm excited about it:

This work is exciting because it shows we can predict phage-bacteria interactions accurately just from genomic data, and we may not actually need thousands of phages/strains to do it. Their phage cocktail recommender system worked great when tested on 100 pathogenic E. coli strains, which is super exciting for phage therapy. This paper also is cool when compared directly to Dimi Boeckaerts’ paper on Klebsiella strain-level prediction (PhageHostLearn), which we covered in a previous Phage Picks issue. That one used a different species and fewer phage-host pairs, a similar but different way of scoring phage killing, and different bioinformatics to pull out the features from the genomes that would be used to train the model. Comparing the two side by side helped me start to internalize some of the decisions you might make when building these models. Overall, I love that it seems each month we get a phage-host prediction model for a new species… soon we’ll have covered them all! And then what? I am excited. ~ Jess

Paper: https://www.nature.com/articles/s41564-024-01832-5 Gaborieau, B., Vaysset, H., Tesson, F. et al. Prediction of strain level phage–host interactions across the Escherichia genus using only genomic information. Nat Microbiol 9, 2847–2861 (2024).

Problem choice and decision trees in science and engineering

What is it about?

Not a phage paper, or even a biology paper, but perhaps even more relevant to us all. This is an excellent commentary by Stanford professor Michael Fischbach tackling an often overlooked aspect of research: how do we choose which problems to work on? He argues that scientists typically rush into execution without spending enough time selecting the right problem to solve (I would agree! We don’t know enough to pick a project, and our PIs want to get us going in the lab as fast as possible — 3 month PhD rotations at US universities do not help with this pressure!).

Dr. Fischbach’s framework for better problem selection is very practical — he suggests spending more time up front on problem choice, using specific techniques he calls "intuition pumps" to generate ideas while avoiding common pitfalls, and evaluating ideas based on both their chance of success and potential impact. He also emphasizes the importance of analyzing your assumptions, developing ideas systematically by fixing one parameter at a time, and being willing to adjust course through what he calls the "altitude dance." Basically doing science on the way you pick what science to do. Genius.

Of note — I found this paper because of Julia Bauman’s fantastic explainer tweet about it (and it’s also a course at Stanford!) — this was BEFORE I realized she actually works in my building!

Why I'm excited about it:

As someone who has always struggled with project design and project scope (how BIG should a project be? For me, for an undergrad student, for a new grad student I’m mentoring? How ambitious is too ambitious for an R01 grant?), this piece really resonates with me. I think it could really help researchers at any career stage make better choices about what problems to tackle. ~ Jess

Paper: https://www.cell.com/cell/fulltext/S0092-8674(24)00304-0 Fischbach, M. A. (2024). Problem choice and decision trees in science and engineering. Nature Chemical Biology, 20, 1-2.

TechBio · Technical Aspects of Phage Research