Journal of Computer-Aided Molecular Design最新文献

A library of 39 amino-benzoxazole derivatives, selected from 57 benzoxazole compounds in the NCI database, was evaluated for their potential as KDR inhibitors using computational docking methods, including CDocker, LibDock, and AutoDock Vina. At a screening concentration of 100 µM, 11 compounds demonstrated over 40% KDR inhibition, with six showing notable activity. The IC50 values of the top six compounds ranged from 6.855 to 50.118 µM, with compound 1 showing the highest inhibitory activity (IC₅₀ = 6.855 µM). Docking studies revealed that compound 1 achieved an AutoDock Vina score of − 7.5 kcal/mol, CDocker energy of − 41.4, and a LibDock score of 140.9 against KDR, indicating strong binding affinity compared with the positive control, sorafenib (AutoDock Vina − 10.7 kcal/mol, CDocker − 43.76, LibDock 96.7). Anti-proliferative assays against A549 and MCF-7 cancer cell lines showed that compounds 16 and 17 were the most effective against A549 cells, achieving inhibition rates of 79.42% and 85.81%, respectively. Compounds 16 and 17 also exhibited the highest activity against MCF-7 cells (IC50 = 6.98, 11.18 µM), respectively. The docking scores for compounds 16 (KDR: Vina − 8.9, CDocker − 32.15, LibDock 105.7) and 17 (KDR: Vina − 11.1, CDocker − 19.15, LibDock 121.9) support their potent interactions with the KDR target. These results suggest that selection of aminobenzoxazole derivatives may serve as promising anticancer agents, potentially through inhibition of KDR, EGFR, and FGFR1 pathways. Future work will focus on optimizing compound 1 to enhance therapeutic efficacy and exploring the roles of EGFR and FGFR1 pathways in the activities of compounds 16 and 17. Additionally, the relatively limited dataset constrained the statistical power for quantitative modeling; we plan to expand the aminobenzoxazole library and develop a validated 3D-QSAR model to visualize pharmacophoric hotspots and guide structure-based lead optimization.

Understanding active functional class (agonist vs antagonist) at the human μ-opioid receptor (μOR) is critical for drug discovery and safety assessment. While recent machine learning models such as ExtraTrees (ET) and message-passing neural networks (MPNNs) achieved ROC AUC scores of 0.915 ± 0.039 and 0.918 ± 0.044, respectively, it remains unclear how target-conditioned interaction features influence functional class detection and how resampling choices (e.g., SMOTE) impact robustness when evaluated under identical, fixed splits. Therefore, we introduce the μOR-Ligand framework—a target-aware view-based hybrid feature selection to improve performance in identifying whether an active ligand is an agonist or antagonist. To realize μOR-Ligand, three views have been constructed: (1) fingerprint, (2) ligand descriptors, and (3) molecular interaction features, yielding a comprehensive feature space of 114,552 variables (1190 fingerprints, 1618 ligand descriptors, 111,741 interaction descriptors). Feature selection is performed per view to obtain three view-specific subsets; each trains a base learner, and their out-of-fold predictions are fused via a linearly weighted multimodel feature selection stage. In parallel, the three selected feature sets are merged and trained with a stacking model (ensemble feature selection). Finally, μOR-Ligand forms a view-based hybrid feature selection by linearly combining the multimodel and ensemble outputs. Such a target-aware view-based hybrid feature selection for the stacked ensembles framework achieved an improved ROC AUC of 0.930 ± 0.026, supported by a promising significant p-value of 0.046 and a t-statistic of 1.707 (> t-critical=1.663) against the recent model, MPNNs. Also, μOR-Ligand further increased ROC AUC to 0.977 on internal cross-validation, as the highest ROC AUC score. In addition, μOR-Ligand is evaluated under a resampling-controlled μOR evaluation protocol that pairs ± SMOTE on identical, fixed splits. Overall, the study (1) demonstrates that target-aware interaction features, though weak alone, contribute a complementary signal in multimodel fusion, improving performance in functional classification and stability, and (2) establishes a resampling-controlled evaluation protocol for μOR modeling, and (3) identifies correlations between top features and μOR pocket chemistry/residues, and (4) case study to show effectiness on unseen external data, as a real-world application. Overall, the study demonstrates that hybridizing ligand-based and target-conditioned views—via target-aware view-based hybrid feature selection for stacked ensembles—adds complementary signal beyond ligand-only baselines, particularly for functional class (agonist vs antagonist).

Zerumbone is a natural sesquiterpene compound from Zingiber zerumbet plant. While it significantly exhibits analgesic properties through the μ-opioid receptor (μOR) found in animal models, its precise molecular mechanism at the receptor level remains poorly investigated. The present work involves 1-µs molecular dynamics (MD) simulations, MM/PBSA binding-free energy analyses, principal component analysis (PCA) as well as Markov state modeling (MSM) to address how the dynamic basis of zerumbone-μOR interactions compared to morphine, which is a known full agonist. MD trajectories reported greater receptor backbone fluctuations, improved loop mobility, reduced stable hydrogen bonds, and moderate receptor compaction in the zerumbone-bound state in contrast to morphine. MM/PBSA calculations indicated similar total binding affinity and the driven for zerumbone affinity was primarily hydrophobic interaction. PCA recognized notable intermediate conformational substates that were stabilized by zerumbone. In the interim, highly stabilized intermediate-activation macrostate with high-kinetic barriers (~ 8–16 k_BT) and millisecond-scale residency was also revealed through MSM analysis. In agreement with analgesic activities reported previously, these computational insights identify zerumbone as a low-efficacy partial agonist, providing comprehensive molecular explanation to its analgesic profile to serve as a source of safer and more opioid-like drugs.

The interaction between staphylococcal protein A (SpA) and human immunoglobulin G (IgG) is pivotal in treating diseases such as cancer, inflammation, infections, and autoimmune disorders. However, acquiring natural SpA variants is labor-intensive, traditional protein design methods often depend on extensive datasets and detailed structural information, limiting their efficiency and applicability. To overcome these limitations, we propose a deep learning-based approach that directly targets desired binding functions by introducing mutations at selected SpA sites to optimize its properties. Specifically, we present a de novo protein design strategy that integrates a diffusion-based generative model with Kidera factor representations to create SpA variants. The framework comprises three modules: sequence generation, where protein sequences are encoded via Kidera factors and novel variants are generated using a diffusion model; computational screening, employing tools like AlphaFold3 to assess structural properties, solubility, and physicochemical characteristics, thereby selecting high-potential candidates; and experimental validation, involving wet-lab experiments to evaluate the biological activities and binding affinities of the designed proteins. The generated SpA variants demonstrated high success rates and strong binding affinities toward IgG. These findings confirm the effectiveness of our method in producing functional proteins comparable to natural counterparts, offering a scalable and data-efficient alternative to protein engineering.

Herein, we report theoretical investigations of the imidazo[1,2-a]pyridine derivative IPD (systematic name 2-(1-(4-bromophenyl)-5-methyl-1H-1,2,3-triazol-4-yl)imidazo[1,2-a]pyridine), and compare the computational outcome with experimental data available from X-ray crystallography studies and spectroscopic analysis. Density functional theory (DFT) was employed as a computational chemistry approach to optimize the geometry and investigate the electronic properties, molecular descriptors, and frontier molecular orbital features of the investigated compound. The DFT-optimized molecular geometry showed good agreement with the experimental structure determined by single-crystal X-ray diffraction (RMSD = 0.2074 Å). The electrostatic potential map of the IPD molecule revealed potential sites for electrophilic attack at the nitrogen in the imidazole ring and at the nitrogen atoms within the 1,2,3-triazole moiety. Additional calculations, however, indicated a higher proton affinity (246.44 kcal/mol) at the aforementioned nitrogen atom in the imidazo[1,2-a]pyridine ring system, suggesting it is the most likely site of protonation. Molecular docking simulations were conducted to investigate the inclusion of the title compound into β-cyclodextrin and to explore the interactions of the IPD molecule with the epidermal growth factor receptor tyrosine kinase (EGFR-TK) as part of an in silico anticancer study. The electronic structures of the docked complexes were further explored using the DFT method, revealing that the intermolecular interactions between the IPD ligand and the receptors also involved a coupling of frontier molecular orbitals.

Alzheimer’s disease (AD) is a neurodegenerative disease with no cure, with aggregates of amyloid-beta (Aβ) plaques, neurofibrillary tangles, and permanent neurodegeneration. Current therapies have been found to provide complementary effects; therefore, there is a need to establish new therapeutic strategies. The neuroprotective activity of Kalanchoe pinnata (KP) was explored in this study using network pharmacology, molecular docking, and in vitro studies. Bioactive compounds with good pharmacokinetic properties have been identified as the 10 bioactive compounds of KP, such as bryotoxin B, kaempferol, and quercetin. A total of 449 common targets of KP and AD that participate in the PI3K-Akt, MAP, and cAMP signaling pathways were identified (AKT1, TNF, and STAT3). Molecular docking results indicated good binding affinities of these KP compounds with AD-related targets. KP aqueous extract (KPAE) inhibited protrophic cytokines and PI3K/Akt signaling in BV-2 microglial cells in a dose-dependent manner by inhibiting Aβ aggregation, antioxidant activity, and neuroinflammation. The above observations indicate that KP has a multi-target effect against AD, which should be proven by preclinical and clinical trials.

Accurate prediction of a drug molecule’s toxicity is a critical step in pharmaceutical research, offering the potential to reduce experimental costs, mitigate adverse effects, and accelerate drug development. Traditional computational methods often rely on handcrafted molecular descriptors, which fall short in capturing the intricate structural and chemical nuances of molecules. In this study, we propose ToxiGraphNet, a graph neural network (GNN)—based regression model for predicting the LD50 value—a quantitative measure of acute toxicity-directly from molecular SMILES strings. Using RDKit, molecules are transformed into graph representations where atoms serve as nodes and bonds as edges, each enriched with chemically meaningful features. Atom features encompass atomic type, degree, aromaticity, chirality, and more, while bond features capture bond type, conjugation, and ring status. These molecular graphs are processed via edge-conditioned convolution layers (NNConv) within the PyTorch Geometric framework, enabling dynamic, chemistry-aware feature aggregation. The model architecture includes three NNConv layers with batch normalization, dropout, and a residual connection to ensure stable training. After global mean pooling, the learned graph-level representations are passed through fully connected layers to predict LD50 values. Training on a curated LD50 dataset yielded impressive performance (MSE: 0.3610, MAE: 0.4424, RMSE: 6009, (R^2): 0.5959), demonstrating strong generalization and predictive accuracy. This work highlights the efficacy of GNNs in modeling molecular toxicity without relying on hand-engineered features and presents a scalable solution for property prediction in drug discovery pipelines.

Graphical abstract

The µ-opioid receptor (µOR) is one of the most important therapeutic targets for drugs worldwide as it plays a fundamental role in pain modulation. Tramadol interacts effectively with µOR for the treatment of pain, with its metabolite (+)-O-desmethyltramadol (M1) being the main cause of its opioid action. However, the structural and pharmacological differences of M1 with respect to opioids do not allow us to fully understand its functioning in the body. In this work, we contribute to the molecular understanding of the mechanism of action of M1. We conduct an exhaustive computational study that integrates molecular docking and molecular dynamics simulations of the µOR-M1 complex. To achieve a comprehensive analysis, we consider eight different conformations for M1, two chair-type and six twisted boat-type. Our study suggests interconversion from twisted boat-type to chair-type conformations and the factors that drive this interconversion, which are, ligand fluctuations, lack of intramolecular bonds, the effect of solvation and conformational energy barriers. We also conclude that to perform protein-ligand molecular modeling it is necessary to use several techniques to achieve reliable results as in our case. These findings contribute to the design of more effective chemical analogues of tramadol.

New porphyrin dyes employing alkoxysilyl anchoring groups were designed using density functional theory (DFT) and time-dependent density functional theory (TD-DFT) for possible use in dye-sensitized solar cells. The new dyes, named SiA series, demonstrated more favorable charge transfer and enhanced light-harvesting efficiency (LHE) compared with SM315 dye, which utilized the conventional cyanoacrylic acid anchoring unit. Among the designed dyes, the SiA-2 displayed the superior light-harvesting properties as shown by the broadened and bathochromically shifted Q-band, most favorable LHE curve, as well as the highest calculated theoretical maximum photocurrent density (({J}_{text{SC}}^{text{max}})). The SiA-2 dye also showcased the most enhanced charge transfer properties based on the calculated transferred charges (({q}^{CT})), charge-transfer distance (({D}^{CT})), and change in dipole moment accompanying intermolecular charge-transfer (({mu }^{CT})). Moreover, the spatial separation distance (r) between the photogenerated hole center and the surface of the TiO₂ semiconductor of SiA-2 suggests a favorable ({V}_{text{OC}}). In this series, SiA-2 emerges as the most promising sensitizer due to its favorable overall characteristics.

Twenty-three 4-anilinoquinazoline derivatives were successfully synthesized, including six new compounds (8, 9, 12, 17, 19, 20) and seventeen known compounds. Seventeen derivatives (10–26) were evaluated for cytotoxic activity against three cancer cell lines (A549, HepG2, and SH-SY5Y) using the MTT assay. The results showed that compound 13 exhibited high selectivity toward the SH-SY5Y cell line with an IC₅₀ value of 13.1 µM, while compound 26 displayed good inhibition against A549 and SH-SY5Y with IC₅₀ values of 24.1 and 14.8 µM, respectively. The ADME analysis further indicated that compounds 13 and 26 possess favorable drug-like and pharmacokinetic properties, supporting their potential suitability for further investigation. Furthermore, molecular docking and molecular dynamics simulations were performed on these two compounds (13 and 26) targeting phosphoglycerate dehydrogenase (PHGDH). The docking results revealed that the fluorine atom exhibited halogen interactions with Tyr173, and the –NH group formed hydrogen bonds with Asp174. Additional hydrogen bond interactions were observed for the nitro group of compound 13 with Gly156 and for the amine group of compound 26 with Leu152. Other interactions were dominated by van der Waals, π–π, π–sigma, alkyl, and π–alkyl contacts with the aromatic N-anilinoquinazoline scaffold. The molecular dynamics simulation demonstrated consistent RMSD, Rg, RMSF, hydrogen bond, and binding energy profiles, confirming the stability and reliability of the PHGDH–ligand complexes in aqueous solution. Notably, compound 13 maintained more persistent hydrogen bonding interactions and induced localized flexibility around the active site compared to compound 26.