Journal of Chemical Information and Modeling 最新文献

{"title":"The Black Hole Strategy: Gravity-Based Representative Sampling for Frugal Graph Learning on Metal-Organic Framework Networks.","authors":"Mehrdad Jalali,A D Dinga Wonanke,Pascal Friederich,Christof Wöll","doi":"10.1021/acs.jcim.5c01518","DOIUrl":"https://doi.org/10.1021/acs.jcim.5c01518","url":null,"abstract":"The expansion of large-scale materials databases has facilitated the development of graph-based representations, encoding structural and functional similarities as edges in data-driven networks. These enable machine learning models to leverage both local features and global relationships. However, densely connected datasets often introduce redundancy and noise, escalating computational complexity without improving performance. Here, we introduce the Black Hole Strategy, a gravity-based representative sampling method that constructs compact, informative subsets from large materials datasets while preserving essential structural and property diversity. Using metal-organic frameworks (MOFs) as a case study, we demonstrate that graph neural networks (GraphSAGE, GCN, and GAT) trained on Black Hole-sparsified datasets achieve comparable or superior classification and regression performance compared to full-dataset models, despite utilizing significantly fewer data points and reduced memory and training time requirements. Analysis of class-level confusion matrices confirms that critical structure-property relationships─such as pore-limiting diameter (PLD)─persist under substantial sparsification. An ablation study on gravity score weights validates the balanced formulation and robustness of the approach. Topological and efficiency benchmarks further demonstrate that the method preserves modularity, diversity, and connectivity across sparsification levels. These findings establish the Black Hole Strategy as a principled and frugal approach for machine learning in materials science, enabling efficient, interpretable, and scalable discovery workflows. Importantly, this work contributes to the objectives of the FAIRmat consortium, which aims to develop a FAIR data infrastructure for condensed matter physics and materials science. Our approach advances FAIR (Findable, Accessible, Interoperable, Reusable) data practices through optimized sampling techniques that enhance data management, reusability, and interoperability in materials informatics.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"33 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145195118","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Modulation of Aβ1-42 Aggregation by a SARS-CoV-2 Protein Fragment.","authors":"Malinda B Premathilaka,Ulrich H E Hansmann","doi":"10.1021/acs.jcim.5c01811","DOIUrl":"https://doi.org/10.1021/acs.jcim.5c01811","url":null,"abstract":"A number of studies have pointed out the possibility that SARS-CoV-2 infections could trigger amyloid diseases such as Parkinson's disease or type II diabetes. In the present study, we probe this question for Alzheimer's disease, which is connected with the presence of amyloids rich in Aβ peptides. For this purpose, we study, by way of molecular dynamics simulations, the interaction between the fragment FKNIDGYFKI of the Spike protein with an Aβ1-42 monomer and two fibril models, one patient-derived and one synthetic. We find that the viral protein fragment appears to shift the ensemble of monomer conformations toward more aggregation-prone ones, and that fibril polymorphs found in patients with Alzheimer's disease appear to be more stabilized than synthetic fibrils. We discuss commonalities and differences in the modulation of amyloid formation by the viral protein fragments by comparing our results with previous studies of other amyloid-forming proteins.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"31 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145195119","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Nanostructured Material Design via a Retrieval-Augmented Generation (RAG) Approach: Bridging Laboratory Practice and Scientific Literature.","authors":"Nikita A Krotkov,Dmitrii A Sbytov,Anna A Chakhoyan,Polina I Kornienko,Anna A Starikova,Maxim G Stepanov,Anastasiia O Piven,Timur A Aliev,Tetiana Orlova,Mushegh S Rafayelyan,Ekaterina V Skorb","doi":"10.1021/acs.jcim.5c01897","DOIUrl":"https://doi.org/10.1021/acs.jcim.5c01897","url":null,"abstract":"The increasing complexity in designing nanostructured materials for electronics, biomedicine, and energy applications requires advanced computational methods to enhance research efficiency and minimize experimental costs. This study proposes an innovative agent-based retrieval-augmented generation (RAG) system integrated with large language models (LLMs) to automate the extraction and analysis of scientific information from extensive literature databases, specifically targeting nanostructured materials developed via two-photon polymerization (2PP). In addition to extracting and analyzing scientific data, our approach emphasizes understanding how these nanostructured materials interact with cells, which is crucial for controlling their application in biomedicine. The developed platform demonstrates robust semantic accuracy (cosine similarity: 0.82) and high overall task precision (0.81), significantly reducing the likelihood of misinformation by incorporating dynamic query refinement mechanisms. The intuitive, user-friendly interface facilitates quick access to relevant scientific data, thereby improving researchers' productivity and enabling more accurate experimental planning. Although the system exhibits certain limitations regarding domain-specific terminology coverage, further fine-tuning and specialized training are anticipated to enhance its performance and reliability for advanced scientific applications.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"6 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2025-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145189457","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Practical Covariance-Based Method for Efficient Detection of Protein-Protein Attractive and Repulsive Interactions in Molecular Dynamics Simulations.","authors":"Mert Golcuk, Mert Gur","doi":"10.1021/acs.jcim.5c01725","DOIUrl":"https://doi.org/10.1021/acs.jcim.5c01725","url":null,"abstract":"<p><p>Molecular dynamics simulations of large protein-protein complexes require scalable analysis. We present a correlation-based workflow that systematically identifies both attractive (stabilizing) interactions, such as salt bridges, and repulsive (destabilizing) interactions, such as same-charge electrostatic repulsions, across extensive interfaces. By constructing an interprotein covariance matrix, filtering residue pairs by distance, and identifying interactions underlying these correlations, our method focuses computational resources on the most relevant regions of the interface while preserving a high level of detail.</p>","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":" ","pages":""},"PeriodicalIF":5.3,"publicationDate":"2025-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145197550","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"In Silico Fold-Switching Protein Design Driven by Cα-Based Statistical Potential.","authors":"Bondeepa Saikia,Anupaul Baruah","doi":"10.1021/acs.jcim.5c01435","DOIUrl":"https://doi.org/10.1021/acs.jcim.5c01435","url":null,"abstract":"Structural plasticity of naturally occurring proteins allows them to change their shape in response to environmental factors such as pH, temperature, or binding partners. This ability to adopt different conformations is essential for many biological processes. While computational methods have been applied to design and redesign protein sequences that fold to a single ordered and stable state, the computational design of protein sequences with high sequence similarity that adopt well-defined but structurally divergent structures remains an outstanding challenge. Here, we designed 28 pairs of sequences using Monte Carlo simulation, denoted as (a1, b1), (a2, b2), (a3, b3), ..., (a28, b28), where ai and bi represent sequences adopting the 3-α fold and 4β + α fold, respectively. Among these, we identified three sets of fold-switching protein sequences, (a1, b1), (a2, b2), and (a3, b3): one with 89.29% sequence similarity and two others with 87.50% sequence similarity. This reflects the ability of statistical potential to finely balance competing structural constraints. The designed sequences differ by only few residues; however, they possess different tertiary structures: a 3-α helix fold and a 4β + α fold. In addition, sequence variants for a1, a2, and a3 are also designed using rational design guided by sequence analysis, and the results show striking outcomes: single point mutations, specifically D26C or A39F in a1, are sufficient to induce fold switching from the 3-α fold to the 4β + α fold while maintaining 98% sequence similarity with the parent sequence. Together, these findings suggest that the design approach is successful in designing fold-switching sequences that are compatible with their respective target structures. This work also ensures that the developed one-body and two-body statistical potentials are successful in designing protein sequences that exhibit fold conservation and the fold-switching phenomenon, as well as stability at the respective target structures.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"102 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2025-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145195093","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MEMO-Stab2: Multi-View Sequence-Based Deep Learning Framework for Predicting Mutation-Induced Stability Changes in Transmembrane Proteins.","authors":"Yihang Bao,Zhe Liu,Hui Jin,Han Wang,Weidi Wang,Guan Ning Lin","doi":"10.1021/acs.jcim.5c01774","DOIUrl":"https://doi.org/10.1021/acs.jcim.5c01774","url":null,"abstract":"Accurately predicting the impact of point mutations on protein thermodynamic stability is essential for understanding structure-function relationships and guiding protein design. This challenge is particularly acute for transmembrane proteins (TMPs), which play vital roles in cellular signaling and drug targeting but remain underrepresented in structural databases. Existing predictors often rely on three-dimensional structures or multiple sequence alignments, limiting their applicability to TMPs due to poor structural coverage and alignment quality. Here, we present MEMO-Stab2, a fast and structure-independent deep learning framework for predicting mutation-induced stability changes in TMPs. MEMO-Stab2 reformulates the task as a binary classification problem, distinguishing destabilizing from neutral mutations based on a ΔΔG threshold of 0.4 kcal/mol. The model integrates multiview features within a Transformer-based architecture, utilizing embeddings from multiple pretrained protein language models (PLMs) and PLM-based structural predictions. By leveraging PLMs, it operates without requiring experimental 3D structures or explicit multiple sequence alignments, implicitly capturing both evolutionary and structural contexts from the amino acid sequence alone. Across internal and external transmembrane mutation data sets, MEMO-Stab2 consistently outperforms existing tools, including specialized predictors and a state-of-the-art general model even after it was fine-tuned on the same domain-specific data, achieving an F1 score of 0.92 on an internal benchmark. Further analyses confirm the model's robustness and specificity. It demonstrates strong generalization across diverse protein families with low sequence identity and shows superior performance in challenging biophysical contexts such as the transmembrane core and interfacial regions. Its validated computational efficiency enables large-scale mutation screening in minutes, offering a practical, robust, and powerful tool for transmembrane protein variant evaluation and engineering.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"19 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145189185","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Molecular Dynamics and Neural Network Analysis Reveal Sequential Gating and Allosteric Communication in FMRFamide-Activated Sodium Channels.","authors":"Jianlin Li,Lu Li,Zhiwei Gao,Yutao Tian","doi":"10.1021/acs.jcim.5c01255","DOIUrl":"https://doi.org/10.1021/acs.jcim.5c01255","url":null,"abstract":"FMRFamide-activated sodium channels (FaNaCs) represent a unique class of neuropeptide-gated ion channels within the degenerin/epithelial sodium channel (DEG/ENaC) superfamily. While cryo-electron microscopy has revealed static binding architectures, the dynamic mechanisms underlying ligand recognition, allosteric signal transmission, and channel gating remain poorly understood. Here, we employed microsecond-scale molecular dynamics simulations coupled with neural relational inference analysis to elucidate the complete activation mechanism of FaNaC at atomic resolution. Our analysis revealed a sophisticated multistage activation process initiated by coordinated dynamics of FaNaC-specific insertions SI1 and SI2. Spontaneous FMRFamide-binding events suggested that SI1 functions as a dynamic gate that facilitates optimal ligand burial and stabilization, while SI2 appeared to serve as a conformational lid stabilizing the bound ligand through thermodynamically favorable induced-fit mechanisms. This ligand-induced conformational change, which involves the cooperative reorganization of the three peripheral loops (L1, L2, and L3) in the extracellular domain, propagates through the extracellular domain, particularly via a coordinated rigid-body motion of the β-ball/palm domain, leading to the reorganization of the central β-sheet in the extracellular vestibule and a subsequent conformational wave that compacts the intracellular vestibule. We further leveraged neural relational inference (NRI) to analyze residue-level allosteric networks, demonstrating that ligand binding enhances the network's connectivity and reorganizes allosteric communication pathways. These findings provide a high-resolution, dynamic view of FaNaC function, revealing a novel gating mechanism for the DEG/ENaC superfamily and laying the foundation for future studies into neuropeptide modulation.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"201 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145189450","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Graph-Based Machine Learning Framework for Predicting Hydrogen Storage Capacity in Metal-Organic Frameworks.","authors":"Azzam Alfarraj,Monther Rashed Alfuraidan,Abdul Malik P Peedikakkal,Ibrahim O Sarumi","doi":"10.1021/acs.jcim.5c01528","DOIUrl":"https://doi.org/10.1021/acs.jcim.5c01528","url":null,"abstract":"Hydrogen is a clean and high-energy fuel, yet its safe and efficient storage remains a key obstacle to widespread adoption. Metal-organic frameworks (MOFs), with their high surface area and tunable porosity, have emerged as promising candidates for solid-state hydrogen storage. In this work, we introduce a graph-based machine learning framework for predicting hydrogen uptake in MOFs by integrating spectral graph theory with data-driven modeling. Molecular structures are represented as weighted graphs from which we extract 20 graph-based descriptors─including Laplacian spectral features, degree statistics, and Zagreb indices─that capture both topological and geometric characteristics of the framework. These interpretable descriptors are used to train multiple regression models on a data set of 3300 MOFs from the Cambridge Structural Database. The XGBoost regressor achieved the highest performance in predicting hydrogen uptake, with a coefficient of determination (R2) of 0.737, RMSE of 0.850% wt, and MAE of 0.433% wt for gravimetric uptake (UG); and a coefficient of determination (R2) of 0.698, RMSE of 4.467 g H2/L, and MAE of 3.045 g H2/L for volumetric uptake (UV). Beyond accurate prediction, the framework enables inverse materials design by identifying graph-based motifs that contribute to improved storage capacity. This integration of chemical graph theory with machine learning provides a scalable, interpretable, and computationally efficient pathway for the discovery of next-generation MOFs tailored for hydrogen storage and other clean energy applications.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"116 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145189452","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Allosteric Mechanisms Triggering Substrate and Cofactor Binding in the SULT1A1 Dimer as Revealed by Molecular Dynamics Simulations.","authors":"Daniel Toth,Balint Dudas,Arnaud B Nicot,Maria A Miteva,Erika Balog","doi":"10.1021/acs.jcim.5c00845","DOIUrl":"https://doi.org/10.1021/acs.jcim.5c00845","url":null,"abstract":"Sulfotransferases (SULTs) are phase II drug-metabolizing enzymes metabolizing a wide range of endogenous compounds and xenobiotics including drugs. SULTs form dimers in vivo, and most isoforms share a conserved dimerization motif. Since it has been shown that the monomers of the SULT1A1 isoform maintain their activity in vitro, the biological significance of dimerization remains unclear. To elucidate the mechanism and the effects of dimerization on the SULT1A1 structure and function, we performed molecular dynamics (MD) simulations on both the monomer and dimer form of the enzyme and investigated the effect of cofactor and substrate binding into the dimer structure and dynamics. Our results show a clear dynamical effect on the dimerization of the apoenzyme, resulting in an increase of the ligand binding gate opening and greater fluctuation of the functional loops of one monomeric subunit. Furthermore, in the dimer, we uncovered intra- and intersubunit allosteric effects as a direct consequence of cofactor and the substrate binding, and we present the corresponding allosteric pathways. Our analyses suggest that the asymmetric behavior of the dimer in the presence of one PAPS molecule may reflect a half-site reactivity mechanism, previously suggested for SULT dimer function, which may be particularly important for large substrates. Thus, our study shed new light in our understanding of SULT1A1 structural dynamics and dimerization as related to enzyme function.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"23 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145189459","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MolDecor: Leveraging Transformers to Decorate Bioactive Molecules for Property Optimization.","authors":"Dibyajyoti Das,Sarveswara Rao Vangala,Arijit Roy","doi":"10.1021/acs.jcim.5c01151","DOIUrl":"https://doi.org/10.1021/acs.jcim.5c01151","url":null,"abstract":"Lead optimization is a critical stage in drug discovery, where promising molecules (lead molecules) are further optimized. It involves the refinement of the chemical structure of the lead molecule to improve its pharmacological properties and drug-like characteristics for development into potential therapies. In this study, we developed a pipeline that includes (a) the creation of a property-specific fragment (decorator) library, (b) learning fragment-scaffold relationship using a BERT-based transformer model, and (c) decorating a given scaffold using fragments from the generated fragment library for improving the properties of the lead molecule. This transformer-based model, MolDecor (Molecule Decorator), was trained on drug-like molecules to learn the optimal decorators for property optimization at single or multiple attachment points on the main scaffold of the lead molecule. The model was fine-tuned on specific property data sets like solubility and affinity using transfer learning to optimize these properties. In this study, an automated method was developed to generate a property-specific decorator library. By learning the relationship between scaffolds and decorators, the model avoids bias toward the most commonly used decorators. This also ensures the easy synthesizability of the generated molecules. The model was tested on the anticancer drug (Thalidomide), an antimalarial molecule (Compound 2), and the estrogen receptor modulator (Cyclofenil) to enhance solubility. Additionally, the model was applied to optimize the affinities of molecules targeting Janus kinase 1.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"31 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145182847","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}