Mustafa Tekpinar

Deep Learning Structural Ensembles as Proxies for Protein Flexibility

This study evaluates deep learning models' ability to predict protein flexibility by comparing their structural ensembles with experimental data, finding AlphaFold3 performs best.

A Collaborative Approach to Improving Missense Mutational Effect Predictions in Oncoproteins

The paper introduces APE score, an ensemble method combining three predictors to improve missense mutation effect predictions in oncoproteins, showing better performance than individual predictors in DMS experiments and providing variant classifications.

PRESCOTT: a population aware, epistatic and structural model accurately predicts missense effect

PRESCOTT predicts the functional impact of missense mutations using sequence and structural data, outperforming existing methods in variant classification and disease-specific mutation analysis.

Comprehensive Mutational Landscape Analysis of Monkeypox Virus Proteome

This study analyzes the mutational landscape of the Monkeypox virus proteome using ESCOTT and provides a resource for assessing new variants.