Understood – thanks for the write-up and engagement, Joel. We will keep the bridge funding situation in mind!

Producing a drug candidate that results in a significant reduction in mortality (idk, like 50% increase in survival in absolute terms?) against at least one virus in vivo, and that shows evidence of efficacy (hard to be specific here) against at least three viruses in vitro, by end of February would be a massive success (noting that how much of a success it is doesn't change appreciably in my eyes with the exact date as long as it doesn't take say a whole year, e.g. I think it would still be a massive success if we took 2 more months than expected)

Not too familiar with what makes a good deadline for prediction markets, but I think we have a pretty good shot of designing + manufacturing candidates by October 6th.

Although noting that designing + manufacturing the candidates (where what constitutes a "design" is pretty unrestricted or just whatever we think has a shot at working, and where manufacturing = making enough for initial tests in cell culture) is relatively straightforward, and I wouldn't consider our ability to do this by a certain deadline as much of an update on anything besides maybe our ability to do labwork at all (if you don't already know us) and how calibrated I specifically am with respect to planning over month-scale timelines.

Getting funding operates on separate timelines, although I think October 6th still works for the purposes of a prediction market.

@joel_bkr Thanks for the update, Joel! I think the references that would be most familiar with my work would be Kyle Fish and Grigory Khimulya from Alvea (Kyle in particular could also talk more about the rest of the team), as well as Kevin Esvelt at MIT. I've sent you their contact info via a DM in Discord. If you do reach out to them, be aware that I won't have let them know to expect outreach from Manifund in particular, although they've provided similar kinds of references for other kinds of opportunities in the past.

@joel_bkr Hey Joel, a few things I’d add/change:

• Regarding why some institutional funders require a higher level of evidence/are more development-focused: I think this is less about expertise (these institutional funders tend to still be highly capable), although that certainly plays a role, and more about how they view their roles within the funding landscape. They see the NIH as having the early-stage stuff “covered”, so they’re trying to influence research beyond that stage. I think there’s also a view that basic science is harder to influence towards particular priorities that an institutional funder might have, since the outcomes of basic science are more uncertain and can be manifold. E.g., if you take the Gates Foundation that has improving global health as a mandate, it would’ve been harder for them to justify funding basic research into mRNA delivery 20 years ago, as there’s not that straight of a line from there to global health (mRNA vaccines may not have worked and could be used for global health just as much as for cancer), and much easier for them to justify funding further development of affordable mRNA vaccines for particular diseases after the basic science was already done. (I think this is more just one factor that drives their view of their position in the funding landscape on the whole than a strong independent view that they would apply on a case-by-case basis, since there’s lots of translational early-stage research where it would be easier to see how they would have outcomes in line with particular funder priorities.)
  
  Insofar as the funding landscape is mostly just “development-focused funding agencies” + “the NIH”, however, this leaves little room for “early-stage science outside of academia” – as I mentioned in a previous reply, NIH reviewers are used to seeing grant proposals that are already the fruits of at least a few years of preliminary data-gathering given the seed funding that academics get. Even with the R21, which is supposed to be an exploratory grant and supposedly requires no preliminary data, in reality heavily favors applicants that already have it (e.g. search Ctrl + F “preliminary data” here to see some frustrated academics). And reviewers would also be more excited to see a principal investigator that more fits the mold of “future superstar academic” – e.g., one at an R1 institution, with first-author Cell/Nature/Science papers in their postdoc, a clear connection between PhD/postdoc research and the proposed research, etc. Even with a collaborative proposal with a well-established investigator, I’d estimate our chances of receiving an NIH R21 at <10% (their base rate is ~15-20%).
  
  Of course, I appreciate the epistemic position that philanthropic funders might be in for evaluating the science, and having to fall back on relying on the same combination of preliminary data + credentials that would be prerequisites for the NIH. It would be a shame though if there were no significant divergence in criteria to get an early-stage science project funded through the NIH or through philanthropic means, and I’m hoping that the combination of a strong impact story + a good reputation in these circles could help pull these apart (or from your perspective, give you an edge in evaluation).

• Regarding what happens if Manifund does not fund PanLabs: I think the likely outcome is not that “the project happens with a significant delay” but rather “the project happens anyway using PanLabs’ existing runway at more risk of not resulting in a strong impact later”, where this risk is a combination of 1) PanLabs is not able to iterate on candidates as much as is necessary to find a good lead candidate, 2) PanLabs has a strong early hit but does not have enough time to capitalize on those good results to secure future funding (e.g., due to length of time to get more grants in, or time to publish), and 3) PanLabs is forced to switch to a primarily for-profit structure due to uncertainty about sustainability of the nonprofit, which takes away the nonprofit arm of the impact story.

@joel_bkr Thanks for the questions – and as a meta-point, it’s been a breath of fresh air to be able to have this real-time, interactive discussion on a funding request, so props to Manifund for enabling this!

Responding to each of your bullet points in order:

• I’m not sure I can cite a base rate for a project like this, as the most relevant category would be “other academic therapeutic-focused projects that yield a compelling in vivo study result”, and I know of many published successes but not the denominator of failures. I expect <50% but don’t know if I can provide much more precision than that. If the project fails, plan B likely involves determining the cause(s) of failure (safety, efficacy, and/or durability) and screening additional drug candidates/formulations that address those. The chances of success are highly contingent on support, i.e. the more iterations we can go through, the higher the chance of success. If we have e.g. a year’s worth of total support, I think our chances of finding a drug candidate that provides durable protection out to 1 month against at least one respiratory virus in mice is 50% (higher than the base rate given relatively less reliance on needing new ideas to pan out, and instead focusing on packaging together existing ideas).

• Yes, although of course depends on the terms, and of course we’d need to incorporate a company first to do this. I do think that the broader vision is better served by the nonprofit, and would be more excited by a world where the nonprofit co-exists in symbiosis with an associated for-profit, rather than one where funding constraints push us into a for-profit as a standalone entity.

• Yes, collaborative grants can help here. We have submitted one collaborative grant application with Japanese academic collaborators already with the GHIT Fund. We could certainly submit another collaborative grant with an academic group that has developed a formulation for e.g. an NIH R21 (small exploratory grant, not as large as an R01). But these grant opportunities tend to be both hypercompetitive and have a ~8-month delay between submission and grant award, so it’s hard for us to make plans with the assumption of receipt of these grants.

• Our experience with the Gates Foundation specifically, who we had engaged at Alvea, is they would require a higher level of preliminary evidence to fund something like this (one of our scientific consultants has hinted at this as well). We are engaged with/have plans to engage with other institutional funders, such as Flu Lab and the DARPA BTO, and we will see how these play out, but I think it could also be the case that these funders see themselves as more development-focused, rather than early-stage research-focused, which they’d rather leave to the NIH. Generally, we think there is a gap in the funding landscape for early-stage research outside of academia (which things like FROs are trying to solve; see here for another take), and we are hoping that philanthropy can help fill in the gap here.

• I’ll let the Alvea executives make their reflections on the closing of Alvea public when they feel it is their time to do so, so I won’t say too much here, other than to say that I would consider the inherent technical risk associated with this project to dominate other factors when considering expectation of success.

• We have not engaged with BARDA at PanLabs, although we did so at Alvea; BARDA is specifically a development authority, which means their remit does not typically cover early-stage research (although I’ve heard rumors they might be trying to broaden their remit soon). E.g., IIRC their arm of Project Next-Gen only funds candidates at Phase 2b of clinical trials onward.

@joel_bkr Good questions! I think the answer depends on if your question is how much money is necessary for PanLabs to 1) stop relying on philanthropic money for further advancement of DNA-encoded DRACOs specifically, or 2) unlock other funding sources for nonprofit sustainability more generally.

For 1), there’s no particular set stage of technology maturity at which for-profit investors would be willing to invest, since there’s a lot of different forms of investment all with different philosophies. For example, it’s plausible that we could apply to an accelerator right now even with no data and be accepted, which would mean launching us down a path that doesn’t require philanthropic funding. Maybe some angels would be willing to invest at this pre-data stage as well. But the more data that we have, and therefore the more de-risked the technology is, the more forms of investment we unlock, so the better chance that we can set up a sustainable for-profit path for the DNA-encoded DRACOs. E.g., we’ve heard the feedback that if we have a compelling animal challenge study (which is the end of Phase 2 in the study outlined in the proposal), that this would be compelling to VC’s. We expect that this kind of data would also probably be necessary to license the IP out to another company for development. So in total, I would say completion of all the studies above (~$300K) is on the higher end of resources required – i.e., the $300K worth of studies would unlock as many opportunities in the for-profit space as possible. Assuming our drug candidate works! If we have to do significant troubleshooting (beyond the troubleshooting built into the plan already), I could see this number doubling.

For 2), I think this investment might need to be more substantial. The analogy to academia is helpful here: a new academic at an R1 institution might typically receive $2 M in start-up funds, which is expected to last them until they can generate enough data to apply for their first NIH R01 grant (the most important one for academics in the biomedical sciences, and also what would be the most relevant source for PanLabs). Academics also wouldn’t have to pay overhead (the university takes care of that instead), graduate student salaries are low, and many students come in on fellowship, so this $2 M could go pretty far. I think we would need to receive at least that much initial funding to start getting NIH funding in, probably moreso given I expect some amount of baseline hesitancy by NIH reviewers to grant to a new, non-educational institution like ours. We might be able to approach other institutional funders like the Gates Foundation sooner than that, but that remains to be seen. So I expect PanLabs as a nonprofit institution would be more reliant on philanthropic funding in the low-single-digit millions for its general broad-spectrum antiviral research activities, before it can fully enter a separate funding environment.

Hey Gavriel, thanks for questions. Will refer you to my reply to Joel regarding the non-profit structure.

As to the minimum, I wasn't sure what the norm was here – I took a look at several other large projects and saw they also set a minimum for $500, so I did the same, given I didn't see a downside. Of course, we appreciate any money that can go towards the project.

If we receive much less than we're asking for, we would make up the difference with runway – we believe the project is worth pursuing even at the risk of further organizational stability. The consequence of this is that we'd be less well set-up to execute the broader PanLabs vision (which I outline in my response to Joel), and this would push us more towards pursuing DNA-encoded DRACOs as a standalone product (assuming the data is positive), probably in a for-profit structure.

Hey Joel! Thanks for your question. There a couple reasons why we decided to be a nonprofit:

1. We experienced the difficulties of Alvea raising in the for-profit sector with technologies that were further along, but that were still deemed not to have enough data to back. Given that we’d be starting at an even earlier stage with PanLabs, we thought that it would be difficult to find anyone to back us in the private sector (noting that biotech investing has dropped substantially over the past couple of years; also noting that our team has a good track record in some respects but none of us have the grey hair and experience shepherding drugs through approval that many investors look for!).

2. PanLabs’ vision goes beyond a single product – our vision is to develop a set of broad-spectrum medical countermeasures that can collectively protect us against all potential pandemic pathogens. Concretely, this means checking each pathogen family that is known to infect humans, and asking ourselves “are we sufficiently protected against this family?” – if not, that spurs additional drug development efforts. A lot of this work will not be commercially relevant (e.g., testing our countermeasures against pathogens that are not widely circulating), so a nonprofit is the most natural home for this vision.

What follows from #1 and #2 is that we see PanLabs as an incubator for ideas for broad-spectrum antivirals, until they are de-risked to the point of being able to go into the for-profit world, either by seeding a new for-profit subsidiary of PanLabs, or through licensing to an external for-profit. This incubation stage would involve doing enough early-stage research until there is enough data and validation for the idea that an investor would feel comfortable backing it. The analogy is to academia – the traditional biotech founding story is “someone develops an idea in academia, and then if that idea is promising enough, licenses that tech out from the university to found a for-profit.” In that story, the for-profit and associated products are treated as a byproduct of curiosity-directed research. PanLabs is hoping to fill the role of academia in incubating new companies and launching new products, except on purpose, and specifically directed towards spinning out medicines for pandemic preparedness.

We view both the nonprofit and for-profit stages of this as necessary to fulfill the mission. The for-profit stage is necessary because only private investment will be able to fund clinical development of a countermeasure, and clinical development is really important to ensure that a drug is actually safe and effective in humans. However, within the for-profit world, you’d only be incentivized to develop the drug against profitable indications, and starting with a single one, so you don’t really get to test breadth particularly against more obscure pathogens. The nonprofit then would be the home to do preclinical testing of countermeasures against a wider panel of pathogens, and would be where we check how close we are to achieving the overall vision. We think there is synergy between the stages – the nonprofit sets up a countermeasure to enter the for-profit stage to begin with, and then extra money coming back from the for-profit (as specified in whatever licensing agreement occurs between PanLabs and the for-profit entity) can sustainably finance the nonprofit’s activities down the line.

However, I will note that we view this as an evolving experiment, and there are a few things that could push us to move into the for-profit sector instead:

• If there is less appetite in the nonprofit sector to support the broader vision than anticipated

• If there is more appetite to fund DNA-encoded DRACOs in the for-profit sector with minimal data than anticipated (e.g., maybe through an accelerator)

• If the first data from the DNA-encoded DRACOs is very positive, such that it makes more sense to go more “all-in” on the DNA-encoded DRACOs rather than broad-spectrum antivirals more broadly

• If we find that there are no other ideas for broad-spectrum antivirals that we are particularly excited about advancing

Sorry for the essay, but I hope that this communicates clearly how we’re thinking about the structure right now!