Rohan David Gnanaolivu

The paper enhances missense variant classification in intrinsically disordered regions by integrating IDR-specific features with existing predictors, significantly improving PR-AUC scores.

This study evaluates how generative AI-generated protein structures improve LOF activity prediction in breast cancer genes compared to experimental structures and other predictors.

The paper functionally characterizes BRCA2 variants using saturation genome editing and classifies them clinically, finding functionally abnormal variants are associated with increased cancer risk.