A multi-scale perspective on the scale-free nature of protein–ligand interaction networks

Abstract

Understanding the intricacies of protein–ligand interactions is fundamental for deciphering molecular mechanisms and developing targeted therapeutics. While high-throughput screening techniques can rapidly identify potential protein–ligand interactions, these screenings typically require considerable resources and time to validate and optimize. Moreover, existing studies on the topological properties of protein–ligand interaction networks often lack comprehensiveness and fail to fully reveal the modular structure and functional diversity of these networks. Our research introduces a multi-scale approach, utilizing experimental data from protein databases to investigate human protein–ligand interactions. We constructed a detailed network of human protein–ligand interactions at the molecular level, which showed how protein–ligand complexes sharing ligands form distinct network modules. At the residue level, we created interaction networks within these network modules and identified and labeled key binding sites in the complexes. This allowed us to determine characteristic differences between binding and non-binding sites. Our study provided new insights into the modularity, redundancy, and flexibility of human protein–ligand interaction networks, offering valuable implications for drug discovery and the identification of new therapeutic targets.