Concordance of Automated ACMG Variant Classification withExpert-Curated Assertions: A Systematic Evaluation Using theClinGen Evidence Repository
This study demonstrates that VarTriage, a streamlined automated classifier utilizing only 10 ClinGen-calibrated criteria and REVEL scores, achieves near-perfect binary concordance (kappa = 0.98) with expert-curated ACMG classifications, though its lower benign sensitivity highlights the need for incorporating non-computational evidence types to further improve accuracy.