A comprehensive DFT study on piroxicam transition metal complexes (Mn, Fe, Co, Ni, and Zn): implications for computational drug design.

Context: Computational approaches are instrumental in understanding the structural and electronic characteristics of drug metal complexes, thereby facilitating the rational deign of more effective pharmaceutical agents prior to experimental validation. The present work was designed to evaluate the drug-likeness and therapeutic potential, bioavailability, pharmacokinetics, and toxicity characteristics of piroxicam transition metal (Mn, Fe, Co, Ni, and Zn). Additionally, the comprehensive structural, electronic, and solvent-dependent behavior were investigated through theoretical analysis, with particular emphasis geometry optimizations and stabilization effects explored in solvents such as water, ethanol, and DMSO. Among the studied systems, the Co and Ni-piroxicam complexes exhibited the highest stabilization energy. HOMO-LUMO energy gaps and molecular electrostatic potential (MEP) maps indicate enhanced charge transfer characteristics, helping to identify reactive electrophilic and nucleophilic sites. These findings underscore the significant influence of solvent polarity and metal ion size on the physicochemical properties and potential bioavailability of metal drug complexes. The ADMET assessment also proved the safety profile coupled with no predicted toxicity, offering meaningful insights for rational drug design.

Methods: ADMET analysis was carried out using the ADMETlab 3.0 online web server to predict, pharmacokinetic behavior, and various toxicity aspects. Furthermore, the geometry optimization and frequency analyses were performed using DFT at the B3LYP/6-31G(d,p) level for non-metal atoms and SDD basis set for transition metals. Solvent effects (water, ethanol, and DMSO) were modeled using the SMD continuum model as implemented in the Gaussian 16. Key descriptors such as HOMO-LUMO energies, their energy gaps, molecular electrostatic potential (ESP), and thermodynamics metrics were computed by Multiwf, Jmol, and SHERMO, respectively.