2.1. Phytochemical Structure The two-dimensional chemical structures of eugenol, cinnamaldehyde, ethyl para-methoxycinnamate, curcumin, hesperidin, levofloxacin, and metronidazole were retrieved from the PubChem Compound Database ( https://pubchem.ncbi.nlm.nih.gov/ accesed on 12 October 2025) of the National Library of Medicine and the National Center for Biotechnology Information [ 22 2.2. Drug Similarity Criteria and ADMET Studies Drug-likeness evaluation was performed for all compounds according to the criteria proposed by Muegge, Egan, Veber, Lipinski, and Ghose. The drug-likeness assessment was conducted using the SwissADME online software ( https://www.swissadme.ch/ accessed on 14 October 2025) by the Swiss Institute of Bioinformatics [ 23]. Pharmacokinetic and toxicity properties (ADMET) were computationally predicted using the pkCSM software, a collaborative project between Instituto Rene Rachou Fiocruz Minas, The University of Melbourne, and the University of Cambridge ( https://biosig.lab.uq.edu.au/pkcsm/prediction accessed on 14 October 2025) [ 24 2.3. Three-Dimensional Structure The amino acid sequences of gyrA, gyrB, and rdxA from H. pylori Indonesian strains were obtained from the Laboratory of the Institute of Tropical Disease (ITD), Airlangga University, Surabaya, Indonesia, and the Faculty of Medicine, Oita University, Yufu, Japan. The sequences were derived from H. pylori populations isolated from gastric biopsy samples of individuals from various regions in Indonesia [ 7]. Ten sequence variants were analyzed for each target protein to account for sequence heterogeneity among Indonesian isolates. The included sequences represented nonredundant isolate variants with complete or near-complete coding regions available for further structural analysis. Each variant was modeled separately and analyzed independently in the docking workflow. Sequence alignment and translation were performed using BioEdit software version 7.7.1 (University of South Carolina, USA) and translated [ 25]. The translated amino acid sequences were submitted to SWISS-MODEL for comparative homology modeling [ 26 Model quality was evaluated using sequence identity, sequence similarity, template coverage, Global Model Quality Estimation (GMQE), QMEANDisCo Global score, Ramachandran plot distribution, MolProbity assessment, Clash Score, rotamer outliers, C-beta deviations, bad bonds, and bad angles. These validation metrics were used to assess the suitability of the predicted models for downstream docking and molecular dynamics simulations. 2.4. Molecular Docking Binding site identification, docking validation, and molecular docking simulations were performed using MVD software (Molegro ପ୍ପ Virtual Docker version 5.5) [ 27]. Docking was performed usign the default MVD search and scoring parameters, except where otherwise specified. The three-dimensional structures of the ligands were built and energy-minimized using using the MMFF94 force field [ 28]. Ligands structures were prepared using the default MVD ligand-preparation workflow. Binding cavities were identified using the cavity search function in MVD. For each target protein, the cavity yielding the most favorable Moldock score was selected as the docking site. The binding site was further validated by re-docking, and the acceptance criterion was established with an RMSD value ≤ 2.0 Å [ 27 Docking simulations were then performed independently for each ligand as each modeled protein target. The Moldock score was used as the primary scoring function, and comparisons were made only within the same docking framework across ligands and targets. Because Moldock score are relative, software-specific outputs, they were interpreted as comparative indicators to evaluate the results of molecular docking. A lower Moldock score indicates more stable binding between the ligand and the receptor, but it is a relative scoring-function output specific to Molegro Virtual Docker and are not directly equivalent to experimentally measured binding free energies. Therefore, comparisons in this study are made primarily within the same scoring framework across ligands and targets [ 29 2.5. Molecular Dynamics Simulation Molecular dynamics simulation using YASARA software ver.25.1.13 [ 30]. In YASARA programming, simulation parameters were configured to approximate physiological conditions of human cells, namely a temperature of 31 °C, a pressure of 1 atm, a pH of 7.4, and a salt level of 0.9% NaCl. Molecular dynamics simulations were conducted for 50 ns (nanoseconds) with trajectory snapshots saved every 25 ps (picoseconds). The Force Field used is AMBER14. 3.1. ADMET Studies The ADMET predictions suggest that the compounds eugenol, cinnamaldehyde, and ethyl para-methoxycinnamate have higher predicted intestinal absorption than levofloxacin and metronidazole, as indicated by their favorable Caco-2 permeability values of greater than 8 × 10 −6 cm/s. Hesperidin exhibits the lowest predicted absorption among the evaluated compounds. The volume of distribution predictions suggest that hesperidin may preferentially distribute into tissues rather than plasma, whereas curcumin appears to be more evenly distributed in plasma. Based on the predicted BBB permeability (logBB), eugenol and cinnamaldehyde may have greater potential to cross the blood–brain barrier (logBB > 0.45 L/kg), whereas curcumin and hesperidin are predicted to have low BBB permeability (logBB 12 Å) and broader fluctuations, suggesting greater conformational variability during the simulation. To complement the RMSD analysis, additional trajectory descriptors including ligand conformational RMSD, radius of gyration, ligand movement, and hydrogen-bond profiles were also evaluated. The radius of gyration remained relatively stable across the simulations, suggesting that the overall compactness of the protein structures was generally maintained throughout the trajectory. The curcumin–gyrA complex additionally demonstrated comparatively lower ligand movement and a more persistent hydrogen-bond interaction profile than the gyrB and rdxA complexes, whereas the latter complexes showed greater fluctuation in ligand mobility and interaction behavior. Taken together, these trajectory-based analyses indicate that curcumin maintained a comparatively more persistent interaction pattern with gyrA than with gyrB or rdxA under the tested simulation conditions. However, the present molecular dynamics analysis remains limited to computational trajectory descriptors, and further analyses such as principal component analysis (PCA), dynamic cross-correlation matrix (DCCM), and free energy landscape (FEL) evaluation may provide additional insight into collective motions and energetically favorable conformational states. 4. Discussion The curcumin–gyrA complex showed comparatively more restrained conformational fluctuation than the curcumin–gyrB and curcumin–rdxA complexes under the same simulation conditions, as indicated by its relatively low and constant total RMSD values. In this context, RMSD is useful for describing overall structural movement during the simulation, but it does not by itself establish biological stability or functional efficacy [ 35]. Therefore, the present MD results should be interpreted as an indication that curcumin maintained a more consistent interaction pattern with gyrA than with the other two targets, rather than as proof of superior antibacterial activity. Since RMSD does not capture local flexibility, hydrogen bonding behavior, or residue-specific rearrangements, the observed trend should be viewed as a preliminary computational signal that requires further structural analysis [ 36 The docking results showed that curcumin and hesperidin produced more favorable Moldock scores than levofloxacin and metronidazole across the evaluated targets (gyrA, gyrB, and rdxA). Within the current docking framework, this suggests that these compounds may interact more favorably with the modeled resistance-related proteins, although the scores should be interpreted only as relative computational estimates rather than direct measures of antibacterial potency. Ethyl para-methoxycinnamate also showed a lower Moldock score than metronidazole for rdxA, indicating a potentially more favorable interaction profile in that specific comparison. Taken together, these findings prioritize curcumin and hesperidin for further experimental evaluation, but they do not demonstrate therapeutic efficacy on their own. The favorable docking profiles observed for curcumin in the present study are consistent with prior reports showing that curcumin and its derivatives can bind bacterial targets with meaningful affinity [ 37]. Curcumin-functionalized chitosan nanosystem (Cur-FCNS) has demonstrated high binding affinity toward several bacterial virulence factors and enhances its activity as an anti- H. pylori agent [ 38]. A novel derivative of curcumin has also been predicted to be a potential competitive ATP inhibitor targeting the catalytic domain of the UDP-N-acetylmuramate-L-alanine ligase (MurC) protein, thereby inhibiting peptidoglycan biosynthesis with the highest predicted binding affinity [ 39]. In the current study, these previous findings support the plausibility of the curcumin–gyrA interaction observed in silico, but they do not substitute for direct validation of the present H. pylori target models. Likewise, the favorable docking profile of hesperidin in this study is in line with earlier reports describing its binding affinity toward other protein targets, although those studies involved different biological systems. In one study investigating its neuroprotective potential, hesperidin exhibited a high binding affinity to the β4 subunit of the Adaptor Protein Complex 4 (AP-4), with a binding energy of −7.2 kcal/mol [ 40]. Another study also demonstrated that hesperidin possesses strong binding affinity toward lipoxygenase, indicating its potential as an antioxidant therapeutic agent [ 41]. Hesperidin has the potential to exhibit significant binding affinity with various bacterial proteins [ 42]. No molecular docking studies have been reported on the interaction of ethyl para-methoxycinnamate with gyrA, gyrB, and rdxA of H. pylori. However, there are studies on cinnamate derivatives as DNA gyrase inhibitors have demonstrated significant binding affinity and antibacterial activity [ 43]. Thus, the present results extend prior work by suggesting that curcumin and hesperidin may warrant further investigation against the modeled gyrA, gyrB, and rdxA proteins of resistant H. pylori strains. These computational findings are supported by several experimental and hybrid in silico studies showing that curcumin and related phytochemicals can exhibit antibacterial activity, improved gastric persistence, and synergistic effects against H. pylori or other bacterial targets. Ejaz et al. (2022) reported that curcumin-functionalized chitosan formulation demonstrated synergistic anti- H. pylori activity in both growth-kinetics and antibiofilm assays, and it performed better than free curcumin and chitosan nanosystems alone [ 38]. Importantly, the formulation also showed a slow cumulative release under simulated gastric conditions, with only 16 ± 0.8% released after 40 h, suggesting that the nanosystem may improve gastric retention and maintain local exposure at the site of infection. Al-Kerm et al. (2023) extended curcumin’s antibacterial relevance by synthesizing a series of curcumin-based heterocyclic derivatives and testing them experimentally [ 37]. Several of the new compounds showed in vitro antibacterial activity, with minimum inhibitory concentrations ranging from 1.56 to 200 µg/mL, and some derivatives displayed additive or synergistic effects when combined with ampicillin. The study also found that one derivative lacked detectable genotoxic effects, while another showed the strongest molecular docking interaction among the synthesized compounds and retained acceptable drug-likeness characteristics. Avgoulas et al. (2021) demonstrated that curcumin-containing system, an oxovanadium(IV)–curcumin complex, showed strong binding to serum albumin, interacted with DNA through a minor-groove binding mode, and exhibited very low hemolytic activity across the tested concentration range, supporting a favorable hemocompatibility profile [ 36]. This suggests that curcumin-based coordination complexes can preserve biologically relevant binding behavior while remaining potentially compatible with clinical use. Therefore, the utilization of structure-based phytochemical screening can be meaningfully connected to laboratory validation when computational predictions are paired with in vitro testing. This approach is supported by the study of Fong et al., who identified oroxindin as exhibiting the strongest antibacterial activity against H. pylori among the screened phytochemicals based on predictive modeling [ 20 The ADMET predictions suggest that curcumin and hesperidin may behave differently from eugenol, cinnamaldehyde, and ethyl para-methoxycinnamate in terms of absorption and distribution. In particular, the lower predicted absorption of curcumin and hesperidin may indicate slower systemic uptake, but this should not be interpreted as reduced relevance for an H. pylori target located in the gastric environment. Hesperidin has showed lower predicted absorption and a greater tendency to remain in tissue compartment than the other evaluated compounds, which may influence its interaction with H. pylori in the gastric environment [ 44]. In addition to its chemical structure, hesperidin has a relatively larger size and molecular weight than the other evaluated compounds [ 45]. Previous studies have also reported that hesperidin exhibits limited intestinal absorption [ 46]. Similarly, curcumin also meets the drug-like similarity criteria according to the results of the ADMET prediction analysis. Previous studies have examined curcumin as a bioactive compound with potential antibacterial relevance across disease models [ 47]. Because the present study is computational, the ADMET results should be viewed as supportive physicochemical information rather than evidence that the compounds will necessarily remain in the gastric mucosa or achieve prolonged local exposure in vivo. Likewise, the predicted toxicity profile is encouraging, but it only indicates the absence of major in silico toxicity signals under the applied model and does not establish human safety. Curcumin and ethyl para-methoxycinnamate meet all the criteria for drug-like similarity. These findings suggest that both compounds possess physicochemical characteristics compatible with commonly used drug-likeness criteria [ 24]. Previous study also reported that curcumin and its derivatives meet the drug-like properties criteria, as Lipinski’s rule of five outlined. This includes molecular docking studies demonstrating their binding affinity to the oncogene protein CagA in H. pylori [ 48]. Other studies have also indicated that curcumin may satisfy Lipinski’s rule of five criteria for drug-like properties. Molecular docking studies of curcumin with the CagA protein from H. pylori demonstrated a stronger binding affinity than amoxicillin and metronidazole [ 49]. Both studies focused on the oncogenic protein CagA and utilized only Lipinski’s rule of five. In contrast, this study employed all relevant drug similarity criteria, including those of Muegge, Egan, Veber, Lipinski, and Ghose, to enhance its validity and reliability. Although in silico studies of ethyl para-methoxycinnamate have not been previously reported, this study provides a foundation for future computational screening research. However, this does not prove that they are drug candidates in the therapeutic sense, but it does indicate that they are reasonable compounds to prioritize in a structure-based screening workflow. In contrast, hesperidin showed substantial deviation from several of the evaluated criteria, which may limit its oral drug-likeness profile despite its favorable docking performance. The present findings therefore indicate that the compounds can be ranked according to their predicted physicochemical suitability and docking behavior, with curcumin emerging as the most consistent computational candidate across the applied analyses. In this study, homology models of gyrA, gyrB, and rdxA were generated for levofloxacin- and metronidazole-resistant Indonesian H. pylori strains. The main value of these models is that they provide a computational framework for comparing phytochemical interactions with resistance-related targets, particularly in a context where experimentally solved structures are limited. To date, the available literature on H. pylori structural models remains limited, especially for gyrA and gyrB, which makes the present homology-modeling workflow a useful exploratory contribution rather than an experimentally verified structural discovery [ 12]. Accordingly, the models should be interpreted as predicted structures that support docking and simulation analyses, with further experimental validation required before any biological inference can be made. Subsequent research may be conducted on the basis of the results of in silico studies, encompassing both in vitro and in vivo [ 50 This study is limited by its fully computational design. The homology models, docking scores, ADMET predictions, and molecular dynamics trajectories provide theoretical evidence of potential target engagement, but they do not constitute experimental proof of antibacterial activity. In addition, the absence of wet-lab validation means that the present results cannot confirm minimum inhibitory activity, enzymatic inhibition, or bacterial growth suppression in H. pylori clinical isolates. Additionally, because the docking analyses were conducted using the standard Molegro Virtual Docker (MVD) workflow and default search/scoring parameters unless otherwise specified, several low-level search parameters (e.g., docking iterations and internal stochastic search settings) were not manually modified or separately archived during the initial computational screening process. Nevertheless, the docking workflow, cavity-selection strategy, RMSD validation threshold, and scoring framework are fully described to support methodological transparency and comparative reproducibility within the same computational environment. The simulations also represent simplified approximations of biological behavior and do not fully capture protein flexibility, host factors, or the complexity of the gastric environment. Therefore, the present findings should be interpreted as a prioritization framework for future experimental work rather than as confirmation of anti- H. pylori efficacy. Homology models of gyrA, gyrB, and rdxA were generated for levofloxacin- and metronidazole-resistant H. pylori Indonesian strains. In the present in silico analysis, ethyl p-methoxycinnamate and hesperidin yield favorable predicted interaction profiles, while curcumin demonstrated the most consistent overall computational performance across the evaluated analysis. Among the three complexes, curcumin–gyrA exhibited the most restrained RMSD profile over time, suggesting comparatively greater conformational persistence than the gyrB and rdxA complexes under the simulation conditions. These findings support the prioritization of curcumin for further laboratory validation as a potential anti- H. pylori agent. Author Contributions Conceptualization, M.G. and M.M.; methodology, M.G. and S.S.; software, M.G.; validation, M.G., S.S., and M.M.; formal analysis, M.G.; investigation, M.G.; resources, M.M.; data curation, M.G.; writing—original draft preparation, M.G.; writing—review and editing, S.S. and M.M.; visualization, M.G.; supervision, M.M.; project administration, M.M. All authors have read and agreed to the published version of the manuscript. Acknowledgments The authors would like to express their sincere gratitude to the leadership of the Faculty of Medicine and Faculty of Pharmacy, Universitas Airlangga, Surabaya, for their institutional support. We also acknowledge the Laboratory of the Institute of Tropical Disease (ITD), Universitas Airlangga, Surabyaa, as well as Faculty of Medicine, Oita University, Yufu, Japan, for providing research facilities and technical support that enabled the completion of this study. All parties included in this section have consented to the acknowledgement. The three-dimensional shape of ( a) gyrA, ( b) gyrB, and ( c) rdxA by SWISS-MODEL prediction. The three-dimensional shape of ( a) gyrA, ( b) gyrB, and ( c) rdxA by SWISS-MODEL prediction. The molecular dynamics simulation trajectory analysis of curcumin complexes with gyrA, gyrB, and rdxA. The figure shows backbone RMSD, ligand conformational RMSD, radius of gyration, ligand movement, total hydrogen bonds, and total RMSD during the simulation period. The molecular dynamics simulation trajectory analysis of curcumin complexes with gyrA, gyrB, and rdxA. The figure shows backbone RMSD, ligand conformational RMSD, radius of gyration, ligand movement, total hydrogen bonds, and total RMSD during the simulation period. The ADMET prediction results. The ADMET prediction results. Biological Activity Eugenol Cinnamaldehyde Curcumin Hespedirin EPMS Levofloxacin Metronidazole Absorption Caco-2 permeability (log Papp in 10 −6 cm/s) 1.559 1.634 −0.093 0.505 1.563 1.365 0.511 Human intestinal absorption (%) 92.041 95.015 82.19 31.481 97.039 97.397 92.759 Distribution VDss (log L/kg) 0.24 0.266 −0.215 0.996 −0.057 −0.028 −0.55 BBB permeability (log BB) 0.374 0.436 −0.562 −1.715 0.111 −0.792 −0.735 Metabolism CYP2DP substrate and inhibition No No No No No No No CYP3A4 substrate and inhibition No No Yes No No No No Excretion Total clearance (log mL/min/kg) 0.282 0.203 −0.002 0.211 0.809 0.414 0.48 Toxicity Max. tolerated dose (log mL/kg/day) 1.024 0.876 0.081 0.525 0.997 0.965 −0.296 Oral rat acute toxicity (LD50) (mL/kg) 2.118 1.88 1.833 2.506 2.127 2.59 1.759 Hepatotoxicity No No No No No Yes No Caco-2, colorectal adenocarcinoma cell line; VDss, volume distribution; BBB, blood–brain barrier; CYP, cytochrome 450; LD50, lethal dose 50. The prediction of drug similarity results. The prediction of drug similarity results. Compound Muegge Egan/Veber Lipinski Ghose XLOGP WLOGP TPSA Rot Bond HBA HBD MLOGP MW WLOGP MR Eugenol 2.27 2.13 29.46 3 2 1 2.01 164.2 2.13 49.06 Cinnamaldehyde 1.9 1.79 17.07 2 1 0 2.01 132.16 1.79 41.54 Curcumin 3.2 3.15 93.06 8 6 2 1.47 368.38 3.15 102.8 Hesperidin −0.14 −1.48 234.29 7 15 8 −3.04 610.56 −1.48 141.41 Ethyl p-methoxycinnamate 3.15 2.16 35.53 5 3 0 2.16 206.24 2.16 58.73 Levofloxacin −0.39 1.2 75.01 2 6 1 0.98 361.37 1.2 101.83 Metronidazole −0.02 0.09 83.87 3 4 1 −0.78 171.15 0.09 43.25 TPSA, topography polar surface area; HBA, hydrogen bond acceptor; hydrogen bond donor; MW, molecular weight; MR, molar refraction. Validation metrics of the predicted H. pylori gyrA, gyrB, and rdxA homology models. Validation metrics of the predicted H. pylori gyrA, gyrB, and rdxA homology models. GyrA Q9ZLD9 P48370 Mean Min Max Template Mean Min Max Template GMQE ୦.୭୬ ବ୍ଦ ୦.୦୧ 0.75 0.76 0.88 ୦.୭୬ ବ୍ଦ ୦.୦୦ 0.75 0.76 0.88 QMEANDisco ୦.୬୯ ବ୍ଦ ୦.୦୫ ୦.୬୮ ବ୍ଦ ୦.୦୫ ୦.୭୦ ବ୍ଦ ୦.୦୫ ୦.୬୯ ବ୍ଦ ୦.୦୫ ୦.୬୮ ବ୍ଦ ୦.୦୫ ୦.୭୦ ବ୍ଦ ୦.୦୫ Seq Identity ୯୫.୯୩ ବ୍ଦ ୦.୦୧ 94.19% 98.67% 100% ୯୫.୩୯ ବ୍ଦ ୦.୦୩ 87.34% 97.95% 100% Seq Similarity ୦.୫୯ ବ୍ଦ ୦.୦୧ 0.58 0.59 0.59 ୦.୫୯ ବ୍ଦ ୦.୦୦ 0.58 0.59 0.6 MolProbity Score ୦.୯୯ ବ୍ଦ ୦.୦୪ 0.93 1.05 0.95 ୦.୯୩ ବ୍ଦ ୦.୦୫ 0.86 1 0.95 Clash Score ୦.୪୩ ବ୍ଦ ୦.୦୫ 0.38 0.53 0.23 ୦.୨୧ ବ୍ଦ ୦.୨୪ 0 0.85 0.23 Ramachandran Favored ୯୫.୧୯ ବ୍ଦ ୦.୦୦ 94.55% 95.79% 95.28% ୯୪.୯୮ ବ୍ଦ ୦.୦୦ 94.55% 95.43% 94.79% Ramachandran Outliers ୧.୦୮ ବ୍ଦ ୦.୦୦୨୮ 0.62% 1.45% 0.73% ୧.୧୬ ବ୍ଦ ୦.୦୦୨୬ 0.50% 1.45% 1.45% Rotamer Outliers ୦.୨୪ ବ୍ଦ ୦.୦୦୧୧ 0.14% 0.42% 1.11% ୦.୮୧ ବ୍ଦ ୦.୦୦୨ 0.28% 0.97% 0.98% C-Beta Deviations ୫.୬ ବ୍ଦ ୧.୭୧ 3 7 4 ୮.୧ ବ୍ଦ ୧.୯୭ 5 11 10 Bad Bonds ୦.୦୦୦୦୬ ବ୍ଦ ୦.୦୦୦୧ 0 0.000302755 0 ୦.୦୦ ବ୍ଦ ୦.୦୧ 0 0.000151561 0 Bad Angles ୦.୦୦୩୫ ବ୍ଦ ୦.୦୦୦୮ 0.002064931 0.004485814 0.004050861 ୦.୪୬ ବ୍ଦ ୦.୦୯ 0.002414903 0.00540054 0.004738802 GyrB Q9ZLX3 P55992 Mean Min Max Template Mean Min Max Template GMQE ୦.୭୩ ବ୍ଦ ୦.୦୦ 0.73 0.73 0.88 ୦.୭୩ ବ୍ଦ ୦.୦୦ 0.73 0.73 0.88 QMEANDisco ୦.୬୫ ବ୍ଦ ୦.୦୫ 0.65 0.65 ୦.୬୫ ବ୍ଦ ୦.୦୫ ୦.୬୫ ବ୍ଦ ୦.୦୫ ୦.୬୫ ବ୍ଦ ୦.୦୫ Seq Identity ୯୮.୧୩ ବ୍ଦ ୦.୦୦ 97% 99% 100% ୯୮.୫ ବ୍ଦ ୦.୦୦ 98% 99% 100% Seq Similarity ୦.୬ ବ୍ଦ ୦.୦୦ 0.6 0.6 0.61 ୦.୬ ବ୍ଦ ୦.୦୦ 0.6 0.6 0.61 MolProbity Score ୦.୫୮ ବ୍ଦ ୦.୦୩ 0.55 0.64 0.65 ୦.୫୯ ବ୍ଦ ୦.୦୫ 0.55 0.73 0.61 Clash Score ୦.୦୭ ବ୍ଦ ୦.୦୪ 0 0.16 0.08 ୦.୦୩ ବ୍ଦ ୦.୦୮ 0 0.14 0.08 Ramachandran Favored ୯୭.୮ ବ୍ଦ ୦.୦୦ 97.41% 97.92% 97.41% ୯୭.୭ ବ୍ଦ ୦.୦୦ 97.67% 97.92% 97.67% Ramachandran Outliers ୦.୨୫ ବ୍ଦ ୦.୦୦୦୪ 0.13% 0.26% 0.26% ୦.୦୫ ବ୍ଦ ୦.୦୦୦୭ 0% 0.13% 0% Rotamer Outliers ୦.୩୦ ବ୍ଦ ୦.୦୦ 0.29% 0.30% 0.44% ୦.୪୬ ବ୍ଦ ୦.୦୦୦୫ 0.44% 0.59% 0.44% C-Beta Deviations ୩.୮ ବ୍ଦ ୦.୭୯ 2 5 2 ୨.୪ ବ୍ଦ ୧.୦୭ 0 3 3 Bad Bonds ୦.୦୦ ବ୍ଦ ୦.୦୦ 0 0.000319693 0 ୦.୦୧ ବ୍ଦ ୦.୦୧ 0 0.000320564 0.000160205 Bad Angles ୦.୩୭ ବ୍ଦ ୦.୦୩ 0.003327787 0.004173125 0.003690037 ୦.୪୦ ବ୍ଦ ୦.୦୪ 0.003566758 0.004753981 0.004408961 RdxA Mean Min Max Template Mean Min Max Template GMQE ୦.୭୬ ବ୍ଦ ୦.୦୦ 0.75 0.76 0.77 QMEANDisco ୦.୭୨ ବ୍ଦ ୦.୦୧ 0.71 0.73 - Seq Identity ୯୬ ବ୍ଦ ୦.୦୧% 94.76% 98.57% 100% Seq Similarity ୦.୬ ବ୍ଦ ୦.୦୧ 0.59 0.61 0.61 MolProbity Score ୧.୭୩ ବ୍ଦ ୦.୧୨ 1.55 1.88 1.59 Clash Score ୩.୦୭ ବ୍ଦ ୧.୧୧ 1.75 4.75 2.18 Ramachandran Favored 91.71 ± 0.00% 90.34% 92.55% 91.59% Ramachandran Outliers 2.16 ± 0.00% 1.20% 3.37% 2.40% Rotamer Outliers 1.47 ± 0.00% 1.05% 1.85% 1.33% C-Beta Deviations ୨.୮ ବ୍ଦ ୧.୬୨ 1 6 1 Bad Bonds ୦.୦୦୦୨ ବ୍ଦ ୦.୦୦୦୪ 0.0000 0.0012 0.0000 Bad Angles ୦.୦୦୬୮ ବ୍ଦ ୦.୦୦୧୫ 0.0049 0.0093 0.0049 Docking cavity coordinates and reference-ligand validation parameters for gyrA, gyrB, and rdxA homology models of Indonesian H. pylori strains. Docking cavity coordinates and reference-ligand validation parameters for gyrA, gyrB, and rdxA homology models of Indonesian H. pylori strains. Strain Volume Surface Area Moldock Score Grid Box RMSD gyrA 745.472 2204.16 −107.176 X −26.23 Y 14.55 Z 51.75 ୦.୧୩ ସ୍ଖ 252.928 788.48 −102.833 104.448 298.24 −89.6367 89.600 313.60 −92.3519 75.264 288.00 −87.1997 gyrB 4413.440 11,928.30 −105.2080 X −15.98 Y 0.76 Z 28.51 ୦.୦୦୨୪ ସ୍ଖ 203.264 573.60 −96.8272 125.952 430.08 −123.0110 44.544 170.24 −79.5600 29.690 102.40 −77.2093 rdxA 355.328 1081.60 −76.3462 X 20.36 Y −9.57 Z 2.6 ୦.୫ ସ୍ଖ 197.632 600.32 −68.4633 171.008 528.64 −83.8158 143.360 433.92 −91.3185 130.560 427.52 −61.9783 Docking was performed using the default MVD search and scoring parameters unless otherwise specified. The reported cavity coordinates correspond to the selected docking regions used for all subsequent ligand-screening analyses. RMSD values were obtained from re-docking validation of the reference ligands within the selected cavities. Moldock scores are relative, software-specific scoring values intended for comparison within the same docking framework and should not be interpreted as absolute binding free energies. Moldock score and root mean square deviation (RMSD). Moldock score and root mean square deviation (RMSD). Compound GyrA GyrB RdxA Moldock Score RMSD Moldock Score RMSD Moldock Score RMSD Eugenol −81.693 + 2.935 0.304 + 0.300 −81.197 + 2.485 0.424 + 0.263 −88.436 + 1.798 0.333 + 1.798 Ethyl p-methoxycinnamate −100.58 + 5.774 0.278 + 0.176 −110.936 + 1.837 0.503 + 0.225 −101.829 + 2.491 0.557 + 2.491 Cinnamaldehyde −73.215 + 4.841 0.189 + 0.117 −73.819 + 0.372 0.182 + 0.118 −79.937 + 3.359 0.232 + 0.230 Curcumin −154.994 + 5.072 0.679 + 0.225 −160.203 + 2.957 0.375 + 0.217 −166.322 + 9.149 0.648 + 0.218 Hesperidin −163.471 + 4.610 0.358 + 0.178 −158.859 + 4.318 0.368 + 0.270 −162.556 + 4.060 0.682 + 0.443 Control −109.158 + 3.030 0.333 + 0.258 −122.587 + 0.566 0.277 + 0.72 −85.610 + 2.054 0.318 + 2.054 RMSD, root mean square deviation.
