AlphaGenome Atlas: in silico mutagenesis of the entire human genome improves prioritization and interpretation of non-coding variants

AlphaGenome predictions of regulatory effects across thousands of molecular phenotypes for every possible single nucleotide variant and many observed indels, condensed into a unified score (AVI) and a genome-wide map of cis-regulatory motifs

Deciphering the functional consequences of non-coding genetic variation remains a major challenge in genomics. AlphaGenome Atlas is a resource enabling the joint interpretation and prioritization of variant effects across the entire human genome: using AlphaGenome, the authors predicted regulatory effects across thousands of molecular phenotypes for every possible single nucleotide variant and many observed indels. These predictions were used to derive a unified, interpretable score, the AlphaGenome Variant Impact (AVI), and to map cis-regulatory motifs across the genome. The authors report that AVI achieved state-of-the-art performance across diverse benchmarks, with improved prioritization of deleterious non-coding variants. Applying the Atlas helped solve an epileptic encephalopathy rare disease case, increased the statistical power to detect rare non-coding variants driving population-level phenotypes and enhanced the mechanistic interpretation of these variants.

Synthesis written by Geno'X. For the full original abstract, please refer to the source publication.

Non-coding variants are already captured by WGS; the challenge is interpreting them, and having predictions for every possible single nucleotide variant, condensed into a single score, addresses that bottleneck directly. The abstract, however, reports no metric: "state-of-the-art performance across diverse benchmarks" allows neither measuring the gain of AVI nor knowing on which data it was established. The clinical reach rests on a single solved epileptic encephalopathy case, in a preprint that has not been peer reviewed.

Analysis by Dr Thibaut Benquey