MISHA

For patrons

One gene, one child, and now a funding problem, not a biology problem.

Early money here is worth the most it will ever be. It de-risks the science at the exact moment it turned, and it buys the infrastructure every future disease branch reuses. Here is the leverage, the model, the ask, and the honest risks.

Who is asking, and why to trust it

I am Misha's father, and I run computational research pipelines for a living. This foundation exists because the standard rare-disease path is too slow for my son's hearing, and because the part of the science that used to be impossible just got cheap. The standing promise: you see the failures next to the wins.

Why a solvable problem stays stuck

The external problem

STRC / DFNB16 has zero approved therapies, and not one lab in Asia working on it. Common enough to matter, rare enough for pharma to skip.

The internal problem

A parent is handed the diagnosis and told, gently, to wait. That helplessness is the real output of the current system.

The deeper problem

Rare diseases die in the gap between a promising result and a fundable program, even when the biology is ready. That gap is exactly what catalytic capital is for.

The plan, in three moves

01

Model it before funding it

Rank the hypotheses and predict the structures in silico, so wet-lab money goes to the best shot, not the first guess.

02

Seed the lab that can move

A 50k to 150k grant buys the preliminary data a lab needs to win the grant that is otherwise out of reach.

03

Recycle the leverage

Each seed that unlocks a large grant, and each tool shared open-access, lowers the cost of the next disease. The infrastructure outlasts the first cure.

The leverage, and why now

The number a venture-philanthropy donor actually cares about: small early capital buys the data that unlocks capital ten to fifty times larger.

10 to 50x

follow-on capital per seed grant

a 50k to 150k grant buys the preliminary data that unlocks the 5M to 50M grant which would not fund without it

And the science just turned. In 2026, a dual-AAV gene therapy restored hearing to near-normal thresholds in a mouse model of this exact gene. This is a funding problem now, not a biology problem.

The leverage compounds past one disease. The pipeline, the registry, and the fundraising model are built once and reused for every case a family ambassador brings. Your capital funds the infrastructure that goes after the next hundred genes.

01

Computation got cheap where it counts

AlphaFold has released 214 million-plus predicted structures and won the 2024 Nobel Prize in Chemistry. The slow part of early structural biology is now hours of inference. Rare disease, being 72% genetic, is the case this helps most.

AlphaFold DB 2024; Nobel Prize 2024; Nguengang Wakap 2020

02

Capital already believes it

Recursion raised a 436M dollar IPO for a rare-disease pipeline. insitro raised 400M at a 2.4B valuation. Isomorphic Labs raised 600M and signed pharma deals worth up to 3B. Computation compressing biology is a funded thesis, not a hope.

Bloomberg 2021; SynBioBeta 2021; TechCrunch 2025

03

Registries are now trial-grade

The FDA accepts natural-history data as an external control arm, which is how tiny populations run rigorous trials. RARE-X already runs a federated patient-data platform across dozens of disease communities. The plumbing exists. Nobody has pointed it at rare hearing genetics from Asia.

FDA 2019; RARE-X

The model has paid off before

Parent-founded, technical foundations have repeatedly turned small early capital into approved therapies. This is the template.

Cystic Fibrosis Foundation

Royalty Pharma, 2014

Invested about 150M dollars over 15 years into the research that became Kalydeco. In 2014 it sold the royalty rights for 3.3 billion dollars and poured it back into the field. Roughly 90% of patients are now eligible for a Vertex therapy, up from about 5%.

Rett Syndrome Research Trust

reverserett.org

Family-funded, roughly 94 cents of every dollar into research, about 92M dollars deployed since 2008. The gene-therapy programs now in Rett clinical trials were incubated at the trust before industry picked them up.

Usher 1F Collaborative

Foundation evolution playbook

Early grants let a Harvard lab pivot into Usher 1F, and that work unlocked a 1.2M dollar follow-on from the Foundation Fighting Blindness. The exact structure we are copying for our first seed grant.

The size of the thing

~300M people with a rare disease worldwide Nguengang Wakap et al., 2020
$997B annual US cost of rare disease (2019) Yang et al., Orphanet J Rare Dis, 2022
$600B+ projected orphan-drug market by mid-2030s Towards Healthcare; Precedence Research
$2.6B average cost to develop one new drug Tufts CSDD, 2014

The business model, on one screen

The whole thing in the Ash Maurya format, read as an impact model rather than a profit one. Unique value in the middle, because that is the part that has to be true.

Problem

More than 7,000 rare diseases, around 300 million people, fewer than 5% with an approved treatment. Each disease is too small for pharma to chase alone, so research stays fragmented. For STRC specifically: zero approved therapies, no lab in Asia.

Customer segments

Beneficiaries: families with rare genetic disease, starting with STRC / DFNB16. Funders: patrons and family offices in Asia, plus grantmakers such as CZI Rare As One, HMRF, and Croucher. Partners: the labs that receive the seed grants.

Unique value

An independent AI lab that de-risks the science with computation before spending wet-lab money, run by a parent with a clock running, publishing everything in public. Shared infrastructure, not one-off charity.

Solution

Four programs. Research Lab (AlphaFold3, hypothesis ranking, open-access). Patient Registry (RARE-X federated, family-controlled). Grants Program (seed grants sized to unlock 10 to 50x follow-on). Families (community, and the honest signal on priorities).

Channels

Warm-introduction patron outreach in year one. The public wiki as the credibility engine. Disease-grant applications from year two. Individual giving once the brand exists.

Revenue model

A Hong Kong company limited by guarantee, Section 88 tax-deductible status pending, plus a US fiscal sponsor (about 5 to 7% fee) so US donors get a receipt. Seed patrons now, grants and individual giving compounding later.

Cost structure

Lean by design. HK incorporation about HKD 4,000 to 7,000. Mandatory annual audit HKD 5,000 to 15,000. Year-one registry 12k to 22k US dollars, near zero on a free platform. First grant 50k to 150k. One founder, no salary before Section 88.

Key metrics

Registry families enrolled. Seed-to-follow-on leverage ratio. Grants disbursed. Open research outputs published. Patron capital committed against target. We publish the real numbers, including the ones that disappoint.

Unfair advantage

Hong Kong as the bridge between Mainland trial pipelines and Western lab quality. Direct lines into the small field of labs moving cochlear gene therapy from mice toward humans. A founder who runs computational pipelines and has urgency no committee can manufacture.

In plain terms: a computationally-run, parent-driven foundation, positioned in Hong Kong, that de-risks rare-hearing-genetics research before it funds it, and publishes everything, including the failures. The first dollars buy the incorporation, the registry, and the first seed grant. Every dollar after that is meant to leave behind a tool the whole field can use.

The ask, and what it buys

Year one targets 100k to 300k US dollars. Concretely, that funds:

  • The HK company and its Section 88 filing, plus the US fiscal-sponsor channel for US donors
  • The RARE-X registry live, with the first families enrolled and IRB approval
  • The first seed grant, 50k to 150k, to a partner lab for STRC mouse-model and vector work
  • The open research outputs, hypothesis ranking and the AlphaFold3 structural set, published on the wiki

Honest note. These targets are ambitious next to the comparable foundations' organic curves. CureSHANK reached about 547k dollars of annual revenue by year three, STXBP1 about 342k, Usher 1F about 157k. We hit our numbers only if a major patron or a large grant lands early. We will publish the real figures, including the gap.

The multi-year plan

Targets, not promises. Budget and headcount scale only with the capital and the milestones.

Horizon
Budget target
Team
Milestone
Year 1
$100k to 300k
1
Incorporated, first patron, registry IRB, first grant
Year 2
$500k to 1.5M
2 to 3
Section 88 status, US channel active, 20+ registry families
Year 3
$2M to 5M
4 to 6
First large grant, 100+ families, mouse-model data
Year 4
$5M to 15M
8 to 12
Asia-native lab begins, China trial-readiness work
Year 5
$15M+
15+
Direct lab operations, multi-target portfolio, pre-IND track

What could go wrong

You should hear it from me.

  • Single-founder key-person risk. Today the foundation is one person wearing every hat.
  • Scientific relationships are early and not yet formalized. We do not name partners until they sign.
  • The year-one to year-three targets are aggressive against the comparable foundations' real curves.
  • Therapy timing is not guaranteed to beat Misha's biological window. That is the honest, disclosed founder bias.
  • Scope risk. The foundation must outlast Misha's specific cure, which is why it spans rare disease broadly, not one variant.
Sources (6)

The cost of waiting is a childhood.

Cochlear development does not pause for funding cycles. If this is yours to fund, the earliest dollars matter most, and they matter now. If you are not ready to write a check, take a twenty-minute call. I answer every email myself.