Structure‐Guided In Silico Design of DprE1 Inhibitors Using Pharmacophore Modeling, Molecular Dynamics Simulations, ADME, and DFT Calculations

Abstract

Decaprenylphosphoryl‐β‐D‐ribose‐2′‐epimerase (DprE1) represents a confirmed target for both covalent and noncovalent antituberculosis agents. In the present study, an integrated computational methodology that involves pharmacophore modeling, virtual screening, pharmacophore‐based design, ensemble docking, molecular dynamics (MD) simulations, MM‐GBSA calculations, density functional theory (DFT), and ADME prediction was utilized to identify potential DprE1 noncovalent binders. A five‐feature DHHRR_1 pharmacophore hypothesis was chosen based on scoring, activity classification, decoy enrichment, and leave‐one‐out robustness parameters. The structural representatives from the available ASINEX screening hits were then used for the rational design of several analogues. Ensemble docking with three DprE1 structures (6HFV, 4P8L, and 4FDO), supported by native‐ligand redocking validation, led to the prioritization of compound 22, which revealed consistently favorable docking profiles in all three receptor conformations. A 200‐ns MD simulation, complemented by PCA, FEL, DCCM, and trajectory‐based MM‐GBSA analyses, supported stable accommodation of compound 22 within the DprE1 binding pocket. DFT analysis provided information about its intrinsic electronic stability, polarization, and reactivity. ADME prediction indicated comparatively favorable predicted solubility and oral absorption, although a potential hERG‐related liability was identified. Overall, compound 22 was computationally prioritized as a promising noncovalent DprE1‐binding candidate requiring further structural optimization and experimental validation.