Matsvei Tsishyn

This study evaluates computational models for predicting pharmacogenomic variant effects, finding that StructureDCA outperforms other methods in distinguishing functional variants.

This paper introduces StructureDCA, a method that improves mutational landscape predictions by integrating structural information into direct coupling analysis.

The paper introduces a simple model using residue conservation and solvent accessibility to predict mutational effects, showing it performs as well as complex deep learning models.

The paper evaluates how evolutionary models predict protein stability using deep mutational scanning data and finds that simple models perform as well as complex ones when combined with structural features.

FiTMuSiC predicts protein fitness effects using structural and coevolutionary data, achieving high accuracy and providing mechanistic insights through functional and structural scores.

This study benchmarks eight predictors of protein-protein binding affinity changes upon mutations, identifying biases and generalizability issues, particularly in machine learning approaches.