Variant predictor trained on functional data beats 48 rivals in clinical tests
A variant effect predictor trained on functional datasets instead of clinical outcomes outperformed 48 existing tools in both functional and clinical benchmarks, according to a study published in Nature Genetics. The tool, FuncVEP, raised accuracy from 78.8% to 84.6% on functional benchmarks and from 90.1% to 92.4% on clinical benchmarks. Researchers used FuncVEP to identify 210 new gene-phenotype associations involving 494 genes linked to inborn errors of immunity in the UK Biobank and the Mount Sinai Million Health Discoveries Program. The discovery rate exceeded that of state-of-the-art predictors.
FuncVEP may improve diagnostic precision for missense variants and accelerate gene discovery in population cohorts by avoiding data circularity inherent in predictors trained on clinical outcomes.
Written by the Genomes desk from the primary source linked above and checked against it. Research use only; not medical advice. Corrections: [email protected].