QSAR-based virtual screening of drugbank entries to identify potent lanosterol 14α-demethylase inhibitors

Background Lanosterol 14α-demethylase (CYP51) plays a critical role in ergosterol biosynthesis and is a well-established target for antifungal drug development. In this study, an integrated computational approach combining ligand-based and structure-based methods was employed to identify novel CYP51 inhibitors. Methods A quantitative structure–activity relationship (QSAR) model was developed using a curated dataset of 222 fluconazole derivatives, comprising 165 active and 57 inactive compounds, with biological activity represented by MIC (Minimum Inhibitory Concentration) and IC50 (Half-Maximal Inhibitory Concentration) values. Molecular descriptors were generated using PaDEL, followed by feature selection and normalization in Orange. Multiple supervised learning algorithms were evaluated, and model performance was validated through 5-fold cross-validation. The validated QSAR model was applied to screen 1337 FDA-approved compounds from DrugBank. Subsequent molecular docking against CYP51 (PDB ID: 5FSA) was performed. The top seven candidates were selected for interaction analysis, and the two most promising compounds underwent 200 ns molecular dynamics simulations. MM/GBSA binding free energy analysis further confirmed the superior binding affinity of Remdesivir (ΔGbind = −66.8 kcal/mol) and Acemetacin (ΔGbind = −48.0 kcal/mol) compared with Fluconazole (ΔGbind = −28.8 kcal/mol), supporting the stability and strength of the predicted CYP51–ligand complexes. Experimental validation through CYP51 inhibition assays and antifungal susceptibility testing is required to confirm the predicted activities of the identified compounds. Results Among the models, the Random Forest classifier achieved the best performance, with an accuracy of 91.5%, precision of 91.4%, specificity of 86.4%, and an auc of 0.934. screening identified 528 compounds as potentially active. molecular docking revealed 216 compounds with binding affinities superior to fluconazole. trajectory analysis demonstrated stable protein–ligand complexes, with remdesivir showing enhanced interaction stability compared to fluconazole. Conclusion Overall, this study highlights the effectiveness of integrating QSAR modeling, molecular docking, molecular dynamics simulations and MM/GBSA for rapid identification of potential CYP51 inhibitors, although further experimental validation and pharmacokinetic studies are necessary to confirm their clinical potential.