Skip to content

UDR Target Explorer

Three scores, never blended: how much evidence, how strong the mechanism, how proven the target.

Ultra Deep Research · target scorecard

How strong is the evidence, really?

One gene, one disease, one card — and three questions kept deliberately apart: how much has been written, how well that writing shows the gene drives the disease, and how strong the evidence is that drugging it helps patients. They disagree often, and the disagreement is usually the finding.

30 gene-disease pairs9 disease areas81 assembled mechanisms5 with a drug in the clinic

Loading…

Evidence Volume is how much has been written about the pair. A high figure means it is well studied; it does not mean the mechanism is proven. Mechanism Strength is how well the assembled evidence shows the gene drives the disease. Target Validation is a different question again — how strong the evidence is that pharmacologically modulating the target produces clinical benefit. A gene can be a poor driver and an excellent target: CD19 is not what causes lupus, it is the handle on the B cells that do. Blending the three would hide exactly the cases worth looking at.

The first two are computed identically for every pair from its own run, so a card reads the same whichever run produced it. Target Validation is curated by hand and carries a source, because the corpus cannot tell a randomised trial from an observational cohort — every entry is marked provisional until a scientist signs it off. Built 17 August 2026; every figure traces back to its run.