Journal of Computer-Aided Molecular Design最新文献

Eukaryotic translation initiation factor 4E (eIF4E) plays a critical role in cap-dependent translation by binding the 7-methylguanosine (m⁷G) cap at the 5′ end of mRNAs, thereby regulating the synthesis of proteins essential for cell growth, survival, and proliferation. Under homeostatic conditions, eIF4E selectively translates a subset of mRNAs; however, in cancer, aberrant signaling leads to persistent activation of eIF4E, promoting tumor progression, metastasis, and resistance to therapy. Despite its clinical relevance, very few studies have explored direct targeting of eIF4E’s cap-binding function using small molecules as a therapeutic strategy. In the present study, we adopted a multi-layered in silico and experimental pipeline to identify small-molecule inhibitors that can effectively disrupt human eIF4E activity. A library of over 400,000 compounds from the ZINC database was virtually screened using the Glide docking protocol in Schrödinger-Maestro. Compounds were shortlisted based on binding affinity and drug-likeness properties. Among the top hits, ZINC145267992, a nucleoside-like molecule, showed promising interaction with the cap-binding pocket of eIF4E. To overcome potential druggability limitations and improve clinical relevance, Entecavir (ETV), a clinically approved antiviral drug for hepatitis B and a structural analogue of ZINC145267992, was identified as a candidate for drug repurposing. Molecular dynamics simulations confirmed the stable interaction of ETV with eIF4E. Our findings not only reinforce the feasibility of targeting eIF4E in cancer but also demonstrate that repurposing FDA-approved drugs like Entecavir could offer a practical and efficient route to therapeutic intervention. This integrative approach opens new avenues for eIF4E-targeted strategies in oncology, aiming to selectively impair oncogenic translation.

The conventional separation of drug-target affinity (DTA) prediction and de novo molecule generation creates a significant bottleneck in drug discovery. To address this, we introduce MutiDTAGen, a unified multi-task learning framework that establishes a bidirectional system between these two complementary tasks. By utilizing shared deep representations and a dynamic optimization strategy, the proposed framework ensures that knowledge from affinity prediction directly guides the generation of target-specific molecules. Our method demonstrates improved performance across multiple benchmarks, achieving, for instance, a 12% reduction in Mean Squared Error (MSE) on the Davis dataset compared to the GraphDTA baseline. This synergistic approach not only enhances prediction accuracy but also improves the quality and target-specificity of generated compounds. By unifying prediction and generation within a single end-to-end architecture, this study offers a unified and efficient computational strategy for drug discovery.

This study elucidates the electronic and structural interplay of 5-(1 H-1,2,4-triazol-1-yl)-2-thiophenecarboxylic acid (TTCA) to assess its potential as a multifunctional heteroaromatic scaffold. Using DFT and TD-DFT calculations at the B3LYP/6-311 + + G(d, p) level, we demonstrate that intramolecular hydrogen bonding locks the triazole and thiophene rings into a highly rigid, planar configuration. This structural coplanarity facilitates extensive π-electron delocalization, which is critical for the molecule’s observed optoelectronic behavior. The analysis reveals a dual electronic character: a chemically stable ground state with a HOMO-LUMO gap of 3.13 eV, contrasted by significant visible-light photoactivity evidenced by a narrow optical transition energy of 1.7 eV. Molecular Electrostatic Potential (MEP) and Non-Covalent Interaction (NCI) analyses identify specific nucleophilic sites and weak interactions that empower TTCA to act as a versatile ligand. Validated by high statistical agreement with experimental literature data for structurally related analogs (FT-IR, NMR, UV–Vis), these results confirm TTCA as a promising candidate for charge-transfer applications, coordination chemistry, and optoelectronic material design.

Three nitro-substituted 2,3-dihydro-1H-perimidine derivatives (ortho, meta-, and para-nitrophenyl) were synthesised via a novel, additive-free ultrasonication-assisted method with high yields (up to 90%). Their structures were validated experimentally and supported by DFT calculations, which also provided insight into the reaction mechanism. Further molecular docking, integrated with MD simulation studies, against the identified EGFR mutants revealed strong binding affinities and stable interactions, especially for the ortho-nitro derivative. To validate these findings, we performed ADME and toxicity analyses that confirmed favourable drug-likeness and safety profiles. Potent anticancer activity consistent with computational predictions was confirmed by MTT assays on NCI-H460 cells. Cell cycle analysis showed that the compounds induced phase-specific arrest, contributing to reduced cell viability. Apoptosis was further validated by Annexin V flow cytometry and AO/EB fluorescence imaging, which revealed early and late apoptotic populations. Overall, the compounds demonstrated strong apoptotic and antiproliferative activity.

Thiosemicarbazones (TSCBZ) are promising insecticidal compounds, but their potential neurotoxicity remains unclear. This study aimed to evaluate the toxicity and cholinesterase inhibition of six substituted TSCBZ derivatives using ligand- and structure-based computational approaches. Electronic property analysis revealed that bromine substitution enhances electrophilicity, while sulfur (in TSCBZ1–3) and nitrogen (in TSCBZ4–6) are the most nucleophilic sites. Toxicity prediction indicated that TSCBZ1, 4, and 6 may induce acute and chronic effects in aquatic organisms. Molecular docking showed that TSCBZ1 and TSCBZ4 exhibit higher affinity for acetylcholinesterase than galantamine, suggesting potential selective inhibition. These findings provide novel insights into the structure–toxicity relationship of TSCBZ and their environmental safety profile.

Pharmacophore modelling is widely used in drug discovery to highlight the key chemical features required for biological activity and to screen large libraries for promising hits. The usefulness of any pharmacophore model, however, depends on how well it is validated. This review brings together the main strategies for assessing pharmacophore model quality, ranging from classical metrics such as ROC-AUC, enrichment factors, and BEDROC to decoy-based evaluations such as DUD-E, as well as visual tools including cumulative gain and lift charts. We also discuss validation workflows built into platforms such as Schrödinger’s Phase module. Each method is described in terms of what it measures, early enrichment, discrimination between actives and decoys, or overall model robustness, and where it is most helpful. By outlining the strengths and limitations of these approaches, this review provides practical guidance for selecting appropriate validation methods and improving the reliability and predictive value of pharmacophore models in virtual screening.

The prevalence of myopia is rising, with genetic and environmental factors playing key roles. Disruptions in circadian melatonin rhythms are also linked to refractive errors, though exact mechanisms remain unclear. Differential expression analysis on the GSE136701 dataset identified circadian-related genes. GSEA, GO, and KEGG analyses revealed key genes. To identify key genes, seven machine learning models were employed, and their performance was evaluated using ROC curve analysis. The correlation genes were further assessed. Concurrently, we performed a comprehensive analysis of immune infiltration and correlation pertaining to these pivotal genes. Additionally, potential target drugs were screened using the DSigDB database, and both protein-protein and molecular docking analyses were performed. To investigate the causal relationship between target genes and myopia, two-sample MR analysis was conducted. Finally, single-cell annotation and cell-cell communication analyses were carried out on the GSE235684 dataset. We preliminarily identified four circadian rhythm-related key genes: UTS2, BTBD9, S100A3, and LGALS9. We identified 8-Bromo-cAMP as the small molecule exhibiting the highest binding efficiency to BTBD9 and UTS2, with a docking binding energy of − 37.5 kcal/mol for the BTBD9-UTS2 complex and − 6.8 kcal/mol for the complex-small molecule interaction. The dynamics simulation analysis has further corroborated the dynamic stability of the UTS2-BTBD9 complex binding with 8-bromo-cAMP. Toxicological profiling of the selected small molecule 8-Bromo-cAMP was conducted using Protox-3.0 and ADMEtlab3.0. Additionally, Mendelian randomization analysis revealed a significant association between UTS2 and myopia, with an inverse variance weighted (IVW) P value of 0.013 (OR 0.532, 95% CI 0.323–0.877). Cochran’s Q test revealed no significant heterogeneity in either MR-Egger (Q = 3.753, P = 0.289) or IVW (Q = 3.893, P = 0.421) estimates, supporting the use of fixed-effects models.

Male reproductive problems mark almost 40% of the total infertility issues worldwide. Other than age, BMI, and different kinds of stress, physical inactivity emerges as one of the prominent causes of infertility in males. This cross-sectional study aimed to investigate the impact of physical activity on male hormones regulated by the hypothalamic-pituitary–gonadal (HPG) axis in 200 healthy adult males residing in Hunza, a hilly area of Pakistan. Out of these, 100 were labelled as sedentary and 100 as non-sedentary based on the absence and presence of physical activity (walking or exercise) in their routine. Male hormones, including testosterone, free testosterone (Free T), follicle-stimulating hormone (FSH), Luteinizing hormone (LH), Sex hormone binding globulin (SHBG), Inhibin, dopamine, and stress markers including Malondialdehyde (MDA), Adenosine deaminase (ADA), and Cortisol were measured by ELISA and compared between the two study groups using Mann Whitney U-test. Pearson correlation was calculated between male hormones and age, BMI, MDA, ADA, and Cortisol. Additionally, four machine learning models: linear regression, ridge regression, random forest, and tuned random forest were developed and compared for prediction modelling focusing on biomarkers (male hormones) based on age, BMI, lifestyle, and stress. Male hormones were observed to be significantly elevated in the non-sedentary males compared to the sedentary ones, except for FSH and LH. In addition, Free T exhibited a strong negative correlation with MDA, ADA, and Cortisol. Of the four predictive machine learning models, tuned random forest proved the most effective with the lowest root mean squared error (RMSE) values for Free T and FSH. The current study findings demonstrated that physically active men had increased levels of reproductive hormones, and free T is strongly inhibited by oxidative, immune, and psychological stress.