Molecular Informatics最新文献

The pharmacokinetic profile of a potential drug is largely determined by its metabolic stability, which reflects its susceptibility to biotransformation. Metabolic stability data allow one to assess the therapeutic value of a compound and its toxicological risk. This assesment relies primarily on pharmacokinetic parameters, particularly half-life (t_1/2) and clearance (CL), which are typically determined using in vitro systems including hepatocytes and liver microsomal fractions. Using the publicly available ChEMBL v. 35 and PubChem databases, we collected over 8000 chemical compounds with experimental intrinsic CL and/or half-life data from liver microsome assays obtained in mice, rats, and humans. Different thresholds were applied to differentiate the stable and unstable molecules. The Naive Bayesian classifier with MNA (Multilevel Neighborhoods of Atoms) descriptors and Self-Consistent Extreme Classifier (SCEC) with QNA (Quantitative Neighborhoods of Atoms) descriptors were used for creating classification models. The accuracy (AUC) of most classification models exceeded 0.85. Self-Consistent Regression was used to create quantitative models. The coefficient of determination of the regression models varied from 0.35 (rat, t_1/2) to 0.7 (human, CL_int). These models were integrated into the freely available web application MetaStab-Analyzer, which provides a unique combination of qualitative (stable/unstable/moderate) and quantitative predictions for three species. A key feature of the application is the providing of numerical metrics for each prediction, which increases its interpretability. This combination of innovative algorithms (SCR and SCEC), dual qualitative-quantitative assessment, and a user-friendly interface is not available in any existing tool. MetaStab-Analyzer is freely available at https://www.way2drug.com/metastab/.

The French Chemoinformatics Society (SFCi) is a learned society that unites French academics, students and industrial scientists in chemoinformatics, promoting this discipline at the interface of chemistry, computer science, data science and AI through conferences, training initiatives, networking and community-building.

Chemical language models (CLMs), particularly encoder-decoder transformers, have advanced generative molecular design. Transformer CLMs are able to learn a variety of molecular mappings for compound design that can be conditioned using context-dependent rules. However, their black-box nature complicates the interpretation of predictions. Current analysis methods mostly focus on attention weights of token relationships or attention flow in encoder and decoder modules and cannot explain predictions at the molecular level. Sequence-based compound design was used as a model system to investigate transformer learning characteristics through systematic control calculations involving modifications of protein sequences and sequence-compound pairs. The analysis revealed that compound reproducibility depended on similarity relationships between training and test data and on compound memorization, while specific sequence information was not learned. These findings indicate that predictions of transformer CLMs are driven by memorization effects and statistical correlations rather than by learning specific chemical or biological information. Understanding this learning behavior aids in avoiding over-interpretation of model outputs and informs the appropriate application of transformer-based CLMs in molecular design.

Staphylococcus aureus is a bacterium classified among the ESKAPE pathogens, which are anticipated to pose a significant global health emergency in the coming decades. The FabI enzyme, present in both Gram-positive and Gram-negative bacteria, is a key enzyme involved in fatty acid synthesis II (FAS-II). In this study, we utilized transformation rules to expand the chemical space from the most potent S. aureus FabI inhibitors. Three newly generated focused libraries, named INDDS, DIADS, and PYRDS, encompassed 172,026 compounds. These compounds were ranked based on structural similarity and predicted pIC₅₀ values obtained from machine learning models. This approach allowed to prioritize compounds in each focused library targeting S. aureus FabI. We analyzed the pharmacological properties and chemical space diversity of the S. aureus FabI inhibitors to gather relevant insights and support the prioritization of compounds for further study. The three newly generated libraries are freely available at https://github.com/DIFACQUIM/S.aureus_inhibitors.

For more than 5 years, several countries in Latin America have been developing and updating compound databases of natural products (NPs) isolated and characterized by their countries. In parallel, multiple research groups have been collaborating and assembling a unified Latin American Natural Product Database (LANaPDB), an open-access compound collection representative of Latin America that stands out as a geographical region distinct from its vastness and richness of NP resources. Herein, we report a significant update of LANaPDB, which gathers NPs from eight countries. Major updates to the database include adding 1,164 new compounds obtained from NaturAr, a NP collection from Argentina published in 2025, and 132 new compounds from Panama. The updated LANaPDB has 14,742 nonduplicate compounds. Moreover, a comprehensive evaluation of 41 ADMET (absorption, distribution, metabolism, excretion, and toxicity)-related parameters was carried out for LANaPDB, and the results were compared with one of the largest NP databases, the Universal Natural Product Database, and the approved small-molecule drugs. The results indicated that the three databases have a very similar ADMET profile. Besides, most of the LANaPDB compounds presented high bioavailability, volume of distribution, plasma protein binding rate, blood-brain barrier penetration, susceptibility to CYP3A4, and half-life less than 12 h. Moreover, most of the LANaPDB compounds were predicted with a low probability of inducing toxicity-related reactions. The third version of LANaPDB and the codes for the curation and determination of 41 ADMET-related parameters are freely available at https://doi.org/10.5281/zenodo.15595030. The code is general and can be used to analyze other compound libraries.

Docking is a structure-based cheminformatics tool broadly employed in early drug discovery. Based on the tridimensional structure of the protein target, docking is used to predict the binding interactions between the protein and a ligand, estimate the corresponding binding affinity, or perform virtual screenings (VSs) to identify new active compounds. This study introduces the ligand B-factor index (LBI), a novel computational metric for prioritizing protein-ligand complexes for docking. Unlike other metrics, LBI directly compares atomic displacements in the ligand and binding site. LBI is defined as the ratio of the median atomic B-factor of the binding site to that of the bound ligand. Using the comparative assessment of scoring functions (CASF-2016) dataset, we evaluated the effectiveness of LBI in guiding the selection of protein-ligand complexes to enhance docking performance. Our results show a moderate correlation (Spearman ρ ~ 0.48) between LBI and the experimental binding affinities, outperforming several docking scoring functions. Additionally, LBI correlates with improved redocking success (root mean square deviation < 2 Å), underlying the significance of a ligand-focused metric. While LBI outperforms other metrics such as the protein B-factor index and resolution, its utility in VS docking remains to be further investigated. LBI is easy to compute, interpretable, applicable in structure-based cheminformatics, and freely available for calculation at https://chembioinf.ro/tool-bi-computing.html.

Ignition cases involving arsons are typically handled by forensic experts who examine spectra of samples collected from scenes of fire to test for the existence or absence of ignitable liquids. This is tedious work, since many cases do not involve such liquids. To facilitate this process, we have developed in this work a Machine Learning (ML)-based workflow for samples' classification based on their gas chromatography (GC) chromatograms (i.e., spectra). To this end, annotated spectra of 181 samples containing three groups of liquids (petroleum distillates, gasoline, and an assortment of other substances) collected from fire scenes as well as two reference databases were obtained from the Israeli Department of Identification and Forensic Sciences (DIFS). These spectra were used for the derivation of ML-based classification models using three algorithms, namely, kNN, representative spectrum, and random forest (RF) giving rise to reliable predictions. To increase the size of the dataset to a level that would enable the usage of more advanced ML algorithms, we have used the experimental spectra to develop a new spectra synthesis algorithm and utilized it to generate a large dataset of synthetic spectra. This dataset was used for the derivation of new kNN, RF, and representative spectrum models as well as deep learning (DL) models producing F1-scores over an independent test set composed entirely of "real" spectra ranging from 0.74-0.95, 0.86-0.95, 0.30-0.75, and 0.85-0.96 for kNN, RF, representative spectrum, and DL, respectively. Following the completion of the work, a second set of real spectra was provided to us by DIFS, and modeling it with the second set of models yielded F1-scores ranging from 0.92-0.96, 0.96-1.00, 0.71-0.82, and 0.95-0.98 for kNN, RF, representative spectrum, and DL, respectively. These results therefore suggest that for this dataset, performances depend more on the size of the dataset used for model training than on the ML algorithm. We propose that the workflow and spectra synthesis algorithm developed in this work could be readily applied to other forensic domains where samples are characterized by spectra, either solely or in combination with other parameters.

In computer-aided drug design, molecular generation models play a crucial role in accelerating the drug development process. Current models mainly fall into two categories: deep learning models with high performance but poor interpretability and heuristic algorithms with better interpretability but limited performance. In this study, we introduce an innovative molecular generation model, the compound construction model (CCMol), which integrates the powerful generative capabilities of the generative pretrained transformer (GPT) and the efficient optimization mechanisms of genetic algorithms (GA) to achieve effective and innovative molecular structures. Specifically, our approach uses structure-based drug design comprising both ligand and protein primary structure-based aspects. CCMol integrates GPT for initial molecular generation and GA for iterative optimization of physicochemical and biological properties. The model's reliability was validated by generating molecules targeting three critical disease-related proteins (GLP1, WRN, and JAK2). The results indicate that CCMol is on average with current advanced models in multiple indicators and performs better than the baseline model in terms of structure diversity and drug-related properties indicators, demonstrating that CCMol exhibits outstanding performance in developing novel and effective candidate drug molecules, particularly suitable for expanding the chemical validity of candidate structures at the early stages of drug discovery.

Lipophilicity is a fundamental physicochemical property widely used to evaluate key parameters in drug design, materials science, and food engineering. It plays a critical role in predicting membrane permeability, absorption, and distribution of compounds. Moreover, lipophilicity is commonly integrated into scoring functions to model biomolecular interactions and serves as an important molecular descriptor in machine learning models for property prediction and compound classification. The election of the appropriate pH-dependent lipophilicity ( $\log D_{\text{pH}}$ ) model is important to ensure its accuracy. The incorporation of the ion apparent partition coefficient ( $P_I^{\text{app}}$ ) into predictions of pH-dependent lipophilicity profiles can be essential for accurately reproducing experimental results. In accordance with the principles for findable, accessible, interoperable, and reusable data to improve data management and sharing, here, we introduce LiProS, a FAIR workflow that is easily accessible through a Google Colab notebook. LiProS assists researchers in efficiently determining the appropriate pH-dependent lipophilicity profile based on the SMILES code of their molecules of interest. In addition, LiProS demonstrated its utility in the analysis of ionizable compounds within the NAPRORE-CR natural products database, enabling the identification of the most appropriate lipophilicity formalism tailored to the physicochemical characteristics of these compounds.