2025/11/17 by Azam Shirali, Shirali, Azam, Vitalii Stebliankin +7
Biochemistry, Genetics and Molecular Biology · Medicine · #Biomolecules (q-bio.BM) #FOS: Biological sciences #Machine Learning in Bioinformatics #Viral Infections and Outbreaks Research #vaccines and immunoinformatics approaches
paper · pdf · doi:10.48550/arxiv.2511.13583
openalex publication_date 2025/11/17 · openalex created_date 2025/11/19 · openalex updated_date 2026/07/28
Protein-protein docking is crucial for understanding how proteins interact. Numerous docking tools have been developed to discover possible conformations of two interacting proteins. However, the reliability and success of these docking tools rely on their scoring function. Accurate and efficient scoring functions are necessary to distinguish between native and non-native docking models to ensure the accuracy of a docking tool. Like in other fields where deep learning methods have been successfully utilized, these methods have also introduced innovative scoring functions. An outstanding tool for scoring and differentiating native-like docking models from non-native or incorrect conformations is called Protein binding Interfaces with Transformer Networks (PIsToN). PIsToN significantly outperforms state-of-the-art scoring functions. Using models of complexes obtained from binding the Ebola Virus Protein VP40 to the host cell's Sec24c protein as an example, we show how to evaluate docking models using PIsToN.