Journal of Cheminformatics最新文献

Lipophilicity is a fundamental physicochemical property that significantly influences various aspects of drug behavior, such as solubility, permeability, metabolism, distribution, protein binding, and excretion. Consequently, accurate prediction of this property is critical for the successful discovery and development of new drug candidates. The classical metric for assessing lipophilicity is logP, defined as the partition coefficient between n-octanol and water at physiological pH 7.4. Recently, graph-based deep learning methods have gained considerable attention and demonstrated strong performance across diverse drug discovery tasks, from molecular property prediction to virtual screening. These models learn informative representations directly from molecular graphs in an end-to-end manner, without the need for handcrafted descriptors. In this work, we propose a logP prediction approach based on a fine-tuned pre-trained GraphormerMapper model, named GraphormerLogP. To evaluate its performance, the model was tested on two datasets: one is compiled by us from publicly available sources and contains 42 006 unique SMILES-logP pairs (named GLP); the second consists of 13 688 molecules and is used for benchmarking purposes. Our comparative analysis against state-of-the-art models (Random Forest, Chemprop, CheMeleon, StructGNN, and Attentive FP) demonstrates that GraphormerLogP consistently achieves competitive or superior predictive accuracy across both datasets, attaining mean absolute error values of 0.251 and 0.269, respectively. The GLP dataset is available in the GitHub repository https://github.com/cimm-kzn/GraphormerLogP/tree/main/data.Scientific contribution This paper presents two key scientific contributions. First, we have collected and carefully curated a large and diverse dataset of molecules with measured logP values, comprising over 42 000 compounds. Second, we propose a Graphormer-based model with a task-specific fine-tuning architecture for logP prediction, tailored to leverage representations learned from reaction data. This model demonstrates high performance in benchmark studies on both established literature data and the newly compiled dataset.

Accurate representation of three-dimensional (3D) molecular structures is essential for quantitative structure-activity relationship (QSAR) modeling; however, it remains unclear whether increasing the level of theory used for quantum-chemical geometry optimization yields a practically meaningful benefit for classical conformation-dependent 3D descriptors (Dragon 3D) and the resulting QSAR performance. Here, we benchmark eight commonly used quantum-chemical (QM) geometry-optimization protocols-from minimal-basis Hartree-Fock (HF/STO-3 G) to def2-based hybrid density functional theory (DFT) and the composite method r$_{}^2$SCAN-3c-across three anticancer activity datasets and ten machine-learning classifiers. Descriptor-level analyses (relative deviation, rank correlation, and chemical-space similarity) reveal systematic method dependence in descriptor magnitudes: high-accuracy def2-based DFT protocols produce highly consistent descriptor spaces, whereas some intermediate/low-level settings introduce larger variability, although molecular rankings remain largely robust (Spearman $\rho >0.95$). In contrast, downstream QSAR performance is only weakly affected by the QM level. Across paired dataset$\times$model blocks ($n=30$), mean balanced accuracies cluster tightly (0.852-0.871). B3LYP/3-21 G achieves the highest overall mean balanced accuracies (BA) (0.8709; 95% CI 0.8565-0.8840), while the lowest mean is observed for HF/STO-3 G (0.8518; 95% CI 0.8371-0.8661); def2-based B3LYP methods are numerically slightly lower ($\sim$0.855-0.856). A repeated-measures omnibus test indicates a statistically detectable method effect (Friedman $p=0.006$) but with a small effect size (Kendall's $W=0.094$), and post-hoc Wilcoxon tests with Holm correction identify only one robust pairwise difference (B3LYP/3-21 G vs. HF/STO-3 G, $p_{\text{Holm}}=0.025$). Thus, the observed performance shifts are marginal in magnitude ($\le$1-2%) compared with the 10-100$\times$ differences in computational cost.To support pragmatic method selection, we propose a two-tier Absolute Efficiency Ratio (AER) framework integrating predictive performance with efficiency and methodological considerations. Overall, these results indicate a non-linear and practically weak relationship between QM geometry-optimization level, classical 3D descriptor fidelity, and QSAR performance, suggesting that QM-level upgrades mainly reshape descriptor values without yielding commensurate or actionable gains in predictive accuracy.

Accurately predicting interactions between drugs is critical for pharmaceutical research and clinical safety. The literature keeps moving toward increasingly complex architectures, yet gains on standard benchmarks are often small. We use a deliberately simple setup that keeps the classifier fixed and swaps only the molecular representation. We compare ECFP4 Morgan fingerprints (MFPs), pretrained graph convolutional networks (GCNs), and MoLFormer embeddings on common DrugBank DDI splits and on an FDA drug-drug affinity benchmark. This design lets us isolate the effect of representation and of split design with minimal confounders.Across leak proof DrugBank splits and the FDA task, MFPs with a shallow head match or surpass much more complex graph and knowledge graph systems while using far fewer parameters. On the Unseen DDI split, MFPs reach AUROC 99.4 and AUPR 98.4, slightly ahead of MoLFormer and the prior state of the art. When we impose a strict scaffold out-of-distribution split, a pretrained GCN leads with AUROC 73.99, which pinpoints when extra capacity helps. On standard splits the absolute gains from added complexity are small. We therefore argue that progress will come mainly from better curated datasets and from rigorous out-of-distribution evaluation, rather than from ever more elaborate models.

Predicting likely sites of metabolism (SOMs), i.e., the atoms in a molecule where metabolic reactions are initiated, is an important component of the computational development pipeline for pharmaceuticals, agrochemicals, and cosmetics. Among SOM prediction tools, FAME3, introduced in 2019, is one of only a few non-commercial models capable of predicting both Phase 1 and Phase 2 SOMs for a wide range of xenobiotics. However, its original implementation posed challenges in maintainability, scalability, and interoperability, which hindered broader adoption. To overcome these limitations, we developed FAME3R, an enhanced version of FAME3 designed to improve computational efficiency and facilitate integration with contemporary cheminformatics workflows. FAME3R introduces several new features, including a novel reliability assessment method based on Shannon entropy and the option to select among various featurization strategies. The tool is available as an open-source Python package, offering both a Python API and a CLI for flexible usage. Additionally, trained FAME3R models can be accessed via a GUI and a REST API hosted on the NERDD web platform.

Topological indices are invariants derived from graphs, and computing these measures for chemical frameworks is found to have a great impact on quantitative structure-activity/property relationships (QSAPR) analysis by providing insights into the structural properties of chemical frameworks. However, calculating and validating these indices for large chemical systems poses computational challenges. To address this, we have developed ChemGraphX, an open-source web tool that efficiently computes various distance and degree-based topological indices, along with entropy measures. ChemGraphX accepts diverse input formats, including .pdb, .mol, adjacency lists, and adjacency matrices, making it versatile for both chemical and graph-theoretical applications. This article presents a brief description of ChemGraphX, a comparative analysis with existing tools, and discussed the capability and validation of ChemGraphX in computational and graph theoretical chemistry.

TheorChem2Blender is designed to connect computational chemistry and 3D modeling without requiring extensive technical expertise from either field, bridging the gap between the two disciplines. Developed in Python, the tool supports common chemical input formats (com, xyz, mol2) to create 3D files for animations, visualization, extended reality media, and 3D printing (glb, fbx, obj, dae, stl). The program allows users to customize atoms and bonds and animate chemical compounds using a variety of representational systems. Pilot testing with expert computational chemists and 3D modelers informed the graphical user interface and cross-platform compatibility. TheorChem2Blender is freely available under the Apache 2.0 license, with comprehensive documentation and tutorials to support adoption by chemists, educators, and 3D artists.Scientific Contribution: TheorChem2Blender is an accessible tool with minimal technical knowledge requirements that bridges the gap between 3D modeling and computational chemistry by providing an interface to convert chemical input formats into 3D files. A key contribution of the tool is its support for customizable structural features, such as atom highlighting, adjustable bond orders, and multiple representational styles. Moreover, the program is the first open‑source tool capable of directly converting xyz and com files into fully manipulable fbx and glb 3D animations. This feature facilitates integration of in-silico simulations into web‑based and extended‑reality pipelines, enabling accurate visualization of chemical structures and transformations.