Validate your target in real humans.
Most drug targets fail because they were never checked against human genetics until it was too late. CohortLayer tests yours against large human cohorts and returns clear, aggregate evidence — replicated across populations, and kept current as the science changes.
For biotech, techbio, and translational teams making high-stakes target decisions.
Grounded in research from
Nature · Nature Genetics · JAMA
Your hypothesis
a target · a variant · a gene–disease link
UK Biobank
tested separately
All of Us
tested separately
Aggregate evidence — replicated
effect size · carriers · does it hold
The wrong target is the most expensive mistake in drug development.
A target that looks right in a model can still fail in humans — and by the time that surfaces in the clinic, the cost is years and hundreds of millions. Human genetic evidence is the strongest early signal that a target is real, which is why getting it right, early, matters more than anything else in the pipeline.
Fewer than 1 in 8 drugs entering clinical trials are ever approved. (Wouters et al., JAMA 2020)
You bring the hypothesis. We return the evidence.
- 01
Target validation
Does modulating this gene change disease risk in humans, and is it safe? You get effect size, carrier counts, and phenotype signals across the cohort. - 02
Cross-cohort replication
We test your hypothesis in two large cohorts separately and report whether it replicates — the strongest filter against a false signal. - 03
Continuous monitoring
When a variant is reclassified or new findings land, we re-run automatically and tell you what moved — so your target's evidence is never stale.
Built for the questions other tools leave open.
Cross-cohort by default
A signal in one population may not hold in another. We replicate across cohorts and merge only the aggregates — so what we report has already survived a second test.
Continuously current
Public evidence updates constantly. We bind it to the cohort on an ongoing basis and flag what changed — not a one-time report you redo in six months.
EU-native, aggregate-only
Genetic data is the most sensitive data there is. We handle it under EU/GDPR-native principles, return aggregates only, and never re-identify or move individual records.
Your hypotheses stay private
Your target is your IP. We test your private hypotheses against the cohorts without exposing them — the evidence comes back to you, and only to you.