Every dossier hands your team 13+ pre-powered, falsifiable hypotheses on a base of hundreds of verified sources. That base is what lets a team take one hypothesis — using public data and simulation, no wet lab — to a trial-design-grade, pre-IND-ready design. To show what's possible, we did it ourselves on the fibrosis dossier. The full worked example is below.
Every dossier hypothesis already ships as a funding-grade scaffold — a real, editable power calculation on labeled assumptions, sitting on a synthesized base of hundreds of verified sources. From that base, a single hypothesis can be matured into a trial-design-grade, pre-IND-ready design: roughly 22 linked analyses across ~19 disciplines, every number reproducible from open, seed-fixed code.
That is the middle rung between what a dossier delivers and what an IND-enabling program does — everything computable before a wet lab, and honest about the three things it deliberately does not do: produce wet-lab data, prove the human efficacy the trial exists to measure, or file an IND. To show what the base makes possible, we built the complete example below on our own fibrosis dossier.
Reframe "does it work?" into "does it beat what the backbone alone does?" — then source that number from real trial evidence. It's the parameter that decides everything.
A biomarker-enrichment design drawn from our own pre-registered public-data analysis — the population a conventional endpoint is structurally blind to.
A translation-risk prior over an unproven effect yields the trial's true probability of success — a defensible number you can take to an investment committee, not a hope.
A 20,000-replication adaptive Monte-Carlo trial with true operating characteristics — power, type-I error, futility — that stops early and cheaply when the effect isn't there.
The exact FDA endpoint-qualification path that makes the trial legal in the first place. Whoever qualifies the readout unlocks every program that comes after.
Causal DAG & ICH-E9(R1) estimand, QSP dose-window, in-silico safety (FDA adverse-event mining), IP/FTO scan, decision analysis, health-economics, and the assembled protocol synopsis, SAP, DSMB charter, and consent a CRO can start from.
We took the fibrosis dossier's Hypothesis 1 — that clearing the persistent myofibroblast is the rate-limiting step for reversal — and carried it to a complete, pre-IND-ready trial design. The full document is below: a grounded control arm, the enrichment population, an honest probability of success, the adaptive design, the FDA endpoint-qualification path, and the assembled protocol.
The actual 15-page document: grounded control arm, biomarker-enrichment population, honest probability of success, the 20,000-replication adaptive design, the FDA endpoint-qualification path, and the assembled protocol, SAP, and DSMB charter. Trial-design-grade, not IND-filed.
Read the full design (PDF)Four figures from the fibrosis trial design. Every one reproduces from the open, seed-fixed code in the deposit.
A Blueprint does not run the trial, prove the biology, or replace the toxicology, chemistry, and manufacturing an IND demands. The efficacy effect is trial-measured, and the enrichment analysis is our own pre-registered, first-party re-analysis. What you get is a decision-grade map of the exact trial the field can't yet run: which patients, which endpoint, which odds, which regulatory move, and which safety line you cross at your peril.
Every dossier carries 13+ hypotheses built to this standard — each one a potential launchpad. Browse the catalog and start from the base.