Graph Neural Network Approaches for Identifying Calpain-10 Inhibitors in Neurological Disorder Therapy

This chapter emphasizes the structural and functional characterization of the calpain receptor via artificial intelligence (AI)-based computational techniques, including protein folding prediction, active site identification, computer-aided drug screening, and molecular dynamics simulations. Ligand interaction experiments were facilitated by the high-confidence structural predictions provided by AlphaFold, serving as a foundation for the AI-generated structure. DeepBindPoc identified regions with potential binding sites, subsequently employing DeepBindGCN and AutoDock Vina for virtual screening. The candidate ligands exhibited acceptable binding constants and stability, as verified by molecular dynamics analysis. Bolazine emerged as the most stable complex, indicating its potential for pharmaceutical applications. AI approaches were effectively employed to find and rank drug-like molecules, offering insights into receptor-ligand interactions and dynamics. Collectively, these findings illustrate the promise of computational methodologies in enhancing drug discovery and structural biology, establishing a basis for future experimental validation and rational drug design.