Computational Screening of Selected Phytochemicals Against Levofloxacin and Metronidazole-Resistant Indonesian H. pylori Strains

The incidence of levofloxacin and metronidazole resistance in H. pylori Indonesian strains is increasing. Conventional approaches to antibacterial discovery are often a protracted process. This study uses structure-based virtual screening to quickly discover anti-H. pylori. This study employed homology modeling, docking, ADMET prediction, and molecular dynamics simulation to evaluate phytochemicals against resistant H. pylori gyrA, gyrB, and rdxA structures from Indonesian strains. Three-dimensional structures were constructed from the amino acid sequences gyrA, gyrB, and rdxA of levofloxacin- and metronidazole-resistant H. pylori Indonesian strains. The results of redocking at the binding sites gyrA (0.13 Å), gyrB (0.0024 Å), and rdxA (0.5 Å) obtained a valid RMSD. Curcumin exhibited the lowest average binding scores across gyrA (−154.994 kcal/mol), gyrB (−159.2033 kcal/mol), and rdxA (−166.322 kcal/mol) compared to other compounds and standard therapies, including levofloxacin (−109.1553 and −122.5873 kcal/mol) and metronidazole (−85.6096 kcal/mol). Molecular dynamics simulation results revealed that the curcumin–gyrA complex exhibited comparatively more restrained fluctuation than the other complexes throughout the simulation, as indicated by a consistently low total RMSD value (4–7 Å). Curcumin demonstrated the most favorable computational interaction profile among the evaluated compounds.