The abstract is read at the publisher; the summary is JClub's.
Full-length human AlphaFold 3 structures improve IRF6 coding-variant classification, raising biophysical AUROC from 0.7844 to 0.8967 and DeltaDeltaG from 0.6461 to 0.8516 in leave-one-gene-out benchmarks.
- Why it matters: Scalar sequence predictors collapse distinct failure modes across the TP63-IRF6-GRHL3 transcriptional network, leaving isoform-specific coding, splice-altering indels, and distal enhancer alleles unresolved for orofacial cleft risk.
- What they did: The study integrated 29 AlphaFold 3 assemblies with thermodynamic DeltaDeltaG calculations and AlphaGenome regulatory modeling across clinical, saturation-mutagenesis, and reporter benchmarks, including 12 zebrafish in vivo assays.
- The result: Stability-versus-interface phase-plane mapping separates VWS1 from PPS alleles (OR = 0.0263, p = 4.652 x 10-3) and distinguishes DNA-contact poisons from dimer-unfolding alleles; AlphaGenome splicing tracks reach AUROC 0.9876.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.