Tag: blockchain

How to Tokenize a Disease.

Drug development is expensive, inefficient, regulated, exploitative and deathly dependent on monopolies.

I propose here an organization and fundraising structure that has the potential to solve the incentive problems around drug development for smaller markets, while also giving more control and potential returns to those contributing both time, money, and energy to all stages of disease research and development.

Current process of Biotech development.

1.Basic Research.
2. Develop a Disease Model.
3. Develop a target.
4. Develop list of potential compounds.
5. Synthesize compounds
6. Develop a panel of in vitro assays
7. (In)Validate molecules on panel
8. In Vivo Analysis of the best ones
9. Optimize lead candidates for production & drugginess
10. Patent Everything
11. Pre-Clinical development
12. Manufacture the compounds in drug form
13. Animal Testing
14. IND Filing
15. Human Testing

This is hyper-simplified and I’d encourage everyone to read Principles of early drug discovery to learn more about the process.

What I’d like to emphasize is that this process is complex, expensive, and takes a long time. Combined, these factors have created an industry where it’s rare for a single party to take a product all the way from idea to market.

Instead, you have a variety of parties all hyper-specialized in one section of the pipeline. Researchers build the disease model, that get’s out-licensed to a small biotech to do discovery, they’ll contract with a CDMO to develop it, and then partner with a larger Pharmaceutical company on clinical testing, who will then handle sales and manufacturing, paying a royalty to everyone upstream.

The two primary reasons a drug fails to make it to market are:

  1. Scientific, for some reason or not the biology just doesn’t work as hypothesized (&),
  2. Economic, the venture fails to secure funding to progress it to the next step.

For the most part, this modular approach works at lowering the risks to a company associated with any one candidate. But increases the risk of any one candidate failing as it changes hands from partner to partner. This risk is compounded when the economic incentives are smaller such as with rare diseases.

If you’re a pharmaceutical company with the resources to take a limited number of candidates through trials are you going to pick:

  • Candidate A: with high odds of success, trials will cost $5M, covers a $25M/yr market, and take 4yrs (or),
  • Candidate B: with avg. odds of success, trials will cost $50M, covers a $2B/yr market, and take 7yrs.

Drug development costs are often quoted as $1B per drug that hits the market; the reality is the majority of the cost is in clinical testing and accounting for failed trials. [Sidenote: It’s a very real possibility that a percentage of failed trials are a direct result of the incentives being more in line with economic bets over scientific support].

Where as the cost to develop a candidate is roughly the same no matter your market size. Clinical testing is proportional to the size of your trial, which is determined by the size of the patient population your drug is targeting. If you’re trying to treat high blood pressure in everyone, that will require a trial with 3,000+ but if only 3,000 people have the disease you’re targeting, your costs will be an order of magnitude less.

Cost = Research + Development + (Clinical Trial x Size of Trial)
Return = Price * Size of Market
ROI = Return / Cost

And while placing smaller bets with better odds seems like the right play for getting more diseases treated. Most biotech returns are in placing large bets on outsized risk for the potential of a big payout. So the biggest cost for a biotech developing a drug for a smaller market is the opportunity cost.

In order to get drugs that have lower or even break even economics developed, we will need to either find efficiencies or improve the incentives around fundraising.

Create a Decentralized Autonomous Drug Development Organization aka Science Guild

I propose creating a Science Guild (DAO) made up of those with the skills and experience necessary to organize requests for proposals (RFPs) relating to the various stages of biotech development. And a token that represents ownership of the disease IP and long term royalty stream.


  1. DAO Creation
  2. Initial Disease Offering & Exchange Creation
  3. Proposal-based Development
  4. Royalty

Phase 1: Create an AragonProject based DAO made up of biotech researchers, MBA’s, entrepreneurs.


Technically there’s no reason you’d have to manage the guild on chain as a DAO on the “blockchain”, but it vastly cuts down on expenses from administration, reporting, and auditing. The purpose of the Guild is not to do the research itself but one of Principal Investigator (PI) i.e. organizing, direction, fundraising and project management.

  • Membership is non-transferable.
  • Must be voted in to the Guild, no vetos.
  • Adding a member requires buy-in or dilution.
  • Membership is necessary for submitting RFPs, creating a disease market, nominating proposals, and
  • Guild votes are democratic.
  • Shares represent ownership of the Guild (and it’s share of any disease)
  • It’s not necessary to be a member of the Guild to participate in discussions, and the majority of the process will take place off Guild.

Phase 2: Create a market with specific disease shares using the token bonding curve method.

The numbers will change depending on the exact inputs, But the goal is to have 33% Underwriter, 33% Speculators, 16% AMM, 16% as rewards for individual researchers or contracted companies that complete milestones.
  1. A member of the Guild will recruit a disease Underwriter. This could be a private foundation, a patient advocacy group, or any wealthy funder.
  2. The Nominating Member will create a proposal scoping the disease, providing as much known information about the disease and the state of treatment, and why it’s a fit for the Guild. As well as a draft RFP for the development of a disease model and target.
  3. The Guild votes to create the market, with the underwriter putting up 25–50% of the funds.
  4. You then do an initial disease offering to raise enough to cover creating a disease model whereby you do a dutch auction for 1M shares. With the underwriting receiving 1 share for every 1 bought by someone else. This ensures a wide range of funders and allows the foundation to double the effect of its funding. Minimal funding should yield Step 1–3 with some redundancy.
  5. 50% of the ETH goes into an operations pool managed by the Guild, along with 500K shares.
  6. 50% of the ETH goes into an Automated MarketMaker (e.g. Uniswap) along with 500K shares. With the liquidity tokens owned by the Guild. But only able to be liquidated into a general operation pool once price hits a certain metric and only above a certain liquidity.
  7. This serves multiple purpose:
    ____1. it allows the Guild to earn money from speculation to fund ongoing operations or further development. Above and beyond what could be achieved from an equivalent $1 of donation.
    ____2. Follow on donors buying shares raises market cap which can be leveraged to propel leads to the next often more expensive stage of development.

Phase 3: Development by Proposals

Cloud labs like Transcriptic enable a lot of the early development and candidate assay work that would previously have required a biotech development organization to exist in a physical location.
  1. Once the disease offering is concluded the Guild members finalize and publish a Request For Proposals (RFP) for basic disease models with bounties attached.
  2. Any researcher or research group may submit a grant proposal if nominated by a member, and member proposal also requires being nominated by another member. The majority of this process like the EIP process will be off-chain where feedback can be given.
  3. After sufficient time, the members vote using cumulative voting.
  4. The top 3 receive development contracts to develop disease models.
  5. The goal being enough data to identify a Target & Activity Panel Design.
  6. The data and IP developed must be owned by the Guild, and how to make that data available in an open way while still being a competitive development company is an active question.
  7. The Guild is in effect an LLC with a legal representative, and these contracts are based in real courts and relationships, not just blockchain smart contracts.
  8. For things that have to touch the legal system there’s been good progress in creating a Limited Liability Digital Association.
  9. On data & research submission, if the Guild votes it as sufficient the submitter receives disease shares they can sell as a bonus or keep for a voice in ongoing development.
  10. Then a review period where anyone may comment or submit their own disease models based on the submitted data.
  11. Once at least one model is approved it opens up the option for people to submit target proposals which includes at least one panel for quantifying activity at that target. All that is needed to submit a proposal is a nomination.
  12. The Guild then votes on its validity to disburse an initial payment to the target proposal, and another for an RFP to setup the assay with a cloud lab like Transcriptic.
  13. Once at least one panel&target is approved, it opens up submissions for hits — Which is the fun part. Anyone may submit as many hits as they want provided they stake some nominal amount to prevent spam, the Guild will then vote (yay or nay on each one).
  14. Conviction voting by Guild (or a very confident Researcher), pays to synthesize and send it to a cloud lab like Transcriptic.
  15. If the hit shows activity the submitter receives a multiple of the required stake + 1 disease share. And the target/assay developer receives a share proportional to the number of hits that convert to leads on their target.
  16. You could then repeat this process with additional assays, until there’s consensus on 1 or more leads being a quality candidate.
  17. Once you have a solid candidate the Guild might want to secure the IP with a patent. To do so you’d just follow the RFP process, with one or more members nominating a law firm along with a quote. Then contract out using Openlaw and pay the firm in their preferred cryptocurrency.

For the remaining steps it’s really best to rely on quality CDMO’s and so you’d use the same RFP & Contract proposal process for everything from picking a lawyer to signing a clinical partnership agreement.

18. Production
19. Processing
20. Fill & Finish
21. Clinical Filing
22. Animal Trial
23. Phase 1
24. Phase 2
25. Phase 3/4
26. Go To Market

Phase 4: Royalty and Burn

The final stage would be the royalty stage, which could be distributed to shareholders by buying and burning shares from the automatic market maker contract (AMM) or simply by depositing funds into the AMM.

The Guild’s ongoing role once a candidate makes it to market would primarily be one of maintenance, ensuring that the ongoing license agreement was held up to and the royalties distributed were accurate. An important note here is the disease shares only represent an economic interest so the guild would have the ability to select a licensee agreement that optimizes for distribution over profit and prevent a hostile takeover. (It’s still an open question if it makes sense to have shareholders vote on everything, only major things or leave it all in the Guilds hand).

Once this final milestone is hit, the Guild would be free (but not required) to cash out the majority of their liquidity pool tokens, using the returns either as a dividend and/or as seed for additional disease offerings.

The Cost of Living

These disease shares would be securities, and that would require technical solutions for automating things like tax accounting on the AMM, quarterly reporting based off the Guilds accounts, and ensuring filter contracts are in place to make sure it’s all compliant. But I think the upfront cost of building these tools would be worth the ability to spin up equity fundraising for research & development projects outside of traditional venture requirements.

And I’ll close this with an example of the problem that inspired this mechanism. Human Insulin. 1.25 Million people in the United States were born with a disease that requires regular injections of insulin in order to not die. The American Diabetes Assoc. collect $150M/yr in donations each year to fund $50M in research. And yet the production and ongoing revenues of a single insulin product Humalog are in the billions, for a drug released in 1996.

Humalog Revenues by LILLY

The discrepancy is that those providing the seed stage support, when it’s just research not only don’t capture any of the financial upside of a successful solution. And if they suffer from the disease, they have no ability to prevent a company like Lilly from raising prices indiscriminately as they did with Humalog. Doubling the price over the last 5yrs, even though it was released in 1996.

Now imagine if the ADA underwrote a disease offering to produce generic Humalog (it’s patent has expired and the IRS allows Program related investments). That’s a $100M endeavor. The ADA could match $25M and you’d then only have to raise $5 from half of those who have been diagnosed with Type 1/2 diabetes(30M in the US). If successful, that venture would be able to provide a patient owned source of insulin to diabetics for $5–10/mo indefinitely.

In Conclusion

The Guild’s mission would to enable those suffering from diseases to leverage capital markets to work for them not exploit them.

Please share feedback and constructive criticism here, or on twitter @jacobshiach.