Predicting Dual-Target Antibacterial Activity and SARS-CoV-2 Docking Affinity of Pendant-Armed Copper (II) Metallodrugs: A Neighborhood Topological and Graph Entropy Approach

Abstract

This study establishes a purely structural, laboratory-free framework for ranking the biomedical efficacy of discrete coordination compounds using advanced chemical graph theory. Focus is placed on six pendant-armed mixed-ligand copper(II) complexes, [CuL  ^(1-3) (diimine)] (1-6), featuring varying degrees of nitro substitution and either 2,2'-bipyridyl or 1,10-phenanthroline co-ligands. By explicitly incorporating the central copper core and its seven coordinate bonds, we map the hydrogen-suppressed molecular graphs and compute eleven neighborhood-degree topological indices alongside their corresponding Shannon graph entropies. Multi-target quantitative structure-property relationship (QSPR) models were validated against experimental antibacterial activity (spanning eight bacterial strains) and molecular docking scores targeting the SARS-CoV-2 main protease. The Augmented Zagreb Index served as an exceptional single-descriptor predictor for mean antibacterial performance (R^2=0.969,F=123.9,p=0.0004,〖 Q〗_LOO^2=0.934), while size-type descriptors 〖(M〗_1,HP) accurately captured the SARS-CoV-2 molecular-docking binding free energies ((R^2≈0.78). These findings demonstrate that neighborhood-sphere structural descriptors bypass intensive quantum-chemical optimization, offering a highly efficient and predictive route for accelerating the discovery and development of heteroleptic metallodrug candidates.