QSAR-guided design, DFT, molecular docking, and ADMET prediction of novel thiazolidinone-based indole derivatives as potent α-glucosidase inhibitors

Background Diabetes mellitus is a global metabolic disorder characterized by postprandial hyperglycemia, posing therapeutic challenges. Inhibiting α-glucosidase is an effective strategy for managing glucose levels. Thiazolidinone-based indole derivatives have potential as α-glucosidase inhibitors, although a comprehensive computational analysis of their structural activity determinants is currently lacking. Objectives This study aimed to develop a validated QSAR model for thiazolidinone-based indole derivatives as α-glucosidase inhibitors, design novel analogs from the most effective parent derivative (compound 5), and assess their pharmacokinetic profiles, binding interactions, and electronic characteristics using an integrated computational approach. Methods A GA-MLR QSAR model was developed using a dataset of 20 thiazolidinone-based indole derivatives with known α-glucosidase inhibitory activity. New chemical analogs were generated from the most potent derivative, docked into the α-glucosidase structure, and subjected to ADMET profiling after reactivity was evaluated using DFT (B3LYP/6–31 G ** ). Results The GA-MLR QSAR model identified three important inhibitory activity descriptors, with predictive performance (R² train = 0.911, Q² cv = 0.832, R² test = 0.697). D3, an analog having a -CF₃ group at the fifth position of the indole ring, showed a predicted IC 50 of 2.72 µM and a docking affinity of −6.2 kcal/mol, while DFT revealed the lowest energy gap (ΔEg = 3.07 eV) and high electrophilicity (ω = 4.23 eV). ADMET profiling predicted no mutagenicity or hERG I inhibition and revealed zero Lipinski violations and adequate gastrointestinal absorption (92.95%); however, D3 and compound 5 were predicted CYP3A4 inhibitors, and all evaluated compounds were predicted hERG II inhibitors. Conclusion D3 represents a computationally promising scaffold with predicted α-glucosidase inhibitory activity warranting experimental validation, including α-glucosidase inhibition assays, cytotoxicity studies, CYP450 microsomal stability, and hERG liability assessment. These findings highlight an integrated computational approach for designing thiazolidinone-based indole derivatives as potential antidiabetic agents.