Matching a patient's nonsense variant to candidate readthrough therapies. Every number here is a model prediction or a measured quantity with a stated scope — not medical advice, not a claim any therapy works, and not a clinical recommendation. A research project, shown for inspection.
Start from a gene or variant and see, for one nonsense variant, how each readthrough therapy is predicted to compare — with what is known and what is not yet modelled shown side by side.
Open →Coverage across the ClinVar nonsense-variant set: which suppressor-tRNA designs reach the most variants, genes and ClinVar conditions, and how far each condition's variants are covered — every figure with its denominator.
Open →The per-variant table: every therapy's predicted readthrough and interval, the suppressor tRNA, the NMD and residue-compatibility flags, and the native-stop safety atlas. A technical view over an example slice.
Open →Predictions rest on ClinVar, MANE and a reproduced baseline efficacy model; the safety atlas rests on public Ribo-seq measured for one drug in one cell line. What is built and what is deliberately not is stated on each page.