Peptriever: a Bi-Encoder approach for large-scale protein–peptide binding search

• DOI: 10.1093/bioinformatics/btae303
• Publication Date: 2024-05-03T12:19:55Z
• Authors: Roni Gurvich; Gal Markel; Ziaurrehman Tanoli; Tomer Meirson
• Keywords: Applications Note; Proteins; Computational Biology; Peptides; Databases, Protein; Biochemistry, cell and molecular biology; Software; Algorithms; Protein Binding

Abstract Motivation Peptide therapeutics hinge on the precise interaction between a tailored peptide and its designated receptor while mitigating interactions with alternate receptors is equally indispensable. Existing methods primarily estimate binding score protein pairs. However, for specific without corresponding protein, it challenging to identify proteins could bind due sheer number of potential candidates. Results We propose transformers-based embedding scheme in this study that can quickly rank millions interacting proteins. Furthermore, proposed approach outperforms existing sequence- structure-based methods, mean AUC-ROC AUC-PR 0.73. Availability implementation Training data, scripts, fine-tuned parameters are available at https://github.com/RoniGurvich/Peptriever. The method linked web application customized prediction https://peptriever.app/.