Small molecule- Protein Interaction by Binding Site Similarity

Abstract

Proteins are essential for numerous biological activities, with their functions often modulated by the binding of small molecules. Understanding protein-ligand interactions is thus vital for gaining insights into protein function and designing novel therapeutic agents. Small molecules bind to specific pockets within target proteins based on their physicochemical properties. The limited diversity of protein shapes allows for identifying analogous binding pockets in other proteins. This project aims to develop a ligand-based tool for predicting binding sites in proteins. The project focuses on characterising the binding sites of proteins interacting with phosphoinositides, a family of phospholipids essential for cellular signalling. These phospholipids vary by the number and location of phosphate groups on the inositol head, recruiting specific proteins to perform distinct functions. The project commences with a comprehensive curation of Protein Data Bank (PDB) files, ensuring a robust foundation for constructing a library identifying interacting residues within the protein binding sites. The relative positioning of these interacting residues is determined through ligand and binding site superimposition. Subsequently, an algorithm is developed to identify binding sites exhibiting similar interacting partner localisation, predicting the specific phosphoinositide likely to bind an unknown site. This study aims to predict potential binding sites in proteins, with the aim of offering a versatile approach for identifying binding sites for various other ligands, ultimately contributing to drug design and enhancing our understanding of biological processes.