Are we selecting the right patients — and are there others we have not considered?

Patient heterogeneity is one of the most common reasons a real biological effect fails to show up in a trial. When a population carries mixed genetic liability for a disease, the signal from a mechanism-selective therapy is diluted. The trial fails. The asset is deprioritised. The biology was real.

5 Prime Sciences identifies the patient subpopulations where genetic liability for the disease of interest is highest, where effect sizes are likely to be largest, and where genetic heterogeneity is most likely to dilute a signal if not accounted for in trial design. This stratification is grounded in genetic evidence, not demographic proxies.

We also identify causal biomarkers — proteomic, metabolomic, and genomic — that are mechanistically linked to the target pathway and track target engagement and disease activity. These are the biomarkers that hold up in validation, support regulatory submissions, and provide payers with the evidence they need to connect mechanism to outcome.

Where the data supports it, we evaluate indication expansion: whether the same pathobiology extends to adjacent diseases or patient populations not currently in scope, and where the human genetic rationale for a label extension is strongest.

For rare disease programs, this framework extends to characterising mutation frequency and distribution across patient cohorts, describing the natural history of disease by genetic subtype, and building the biomarker evidence base that regulatory agencies expect to see.

This work has contributed to a $190M raise and Phase 2 advancement for a partner, and has shaped program advancement, prioritisation, and early termination decisions across 19 pharma, biotech, and VC partnerships.