Journal of Chemical Information and Modeling 最新文献

{"title":"PDI-Spot: A Robust Representation Learning Framework\u0000for Identifying Energetic Hotspots in Protein–DNA Complexes","authors":"Esra Basaran,Attila Gursoy,Ozlem Keskin,Nurcan Tuncbag","doi":"10.1021/acs.jcim.6c01579","DOIUrl":"https://doi.org/10.1021/acs.jcim.6c01579","url":null,"abstract":"Abstract The binding affinity of protein–DNA complexes is concentrated within a subset of interface residues, called “hotspots”, which are critical for maintaining stable interactions. Therefore, identifying these energetic anchors is essential for elaborating on how mutations disrupt transcriptional regulation and for identifying druggable regions within disease-associated interfaces. In this study, we introduce our novel prediction model, PDI-Spot, that leverages representation learning to capture the complex features of protein–DNA interfaces. Unlike traditional methods that rely on limited, handcrafted descriptors, PDI-Spot extracts high-dimensional, structure- and sequence-driven representations to achieve better predictive performance. We trained the PDI-Spot model on a curated protein–DNA complex data set and rigorously validated its performance on an independent test set to ensure generalizability across different structural families. Following the extraction of residue-level embeddings, an optimal feature set was identified through a combined approach of minimum redundancy maximum relevance and multilayer perceptron architectures. By training a neural network on the optimal feature set, we developed a robust predictive framework for hotspot identification. On an independent test set, PDI-Spot achieved an AUROC of 0.809, an F1-score of 0.667, and a Matthews correlation coefficient of 0.625. We also verified via permutation tests that PDI-Spot captures biologically meaningful patterns beyond random chance. PDI-Spot consistently demonstrates competitive performance compared to state-of-the-art methods, providing a reliable model for mapping the energetic landscapes of protein–DNA interfaces. The source code and pretrained models are publicly available at https://github.com/ebsrn0/pdi-spot.git.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"17 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149210146","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":"TMC-LewisGen: Automated\u0000Generation of Lewis-like Structures\u0000for Transition–Metal Complexes","authors":"Farshad Shiri,Jingyuan Zhu,Sin Hang Tong,Liren Xiao,Zhenyang Lin","doi":"10.1021/acs.jcim.6c02278","DOIUrl":"https://doi.org/10.1021/acs.jcim.6c02278","url":null,"abstract":"Abstract Lewis structures constitute a foundational language in chemical education, mechanistic reasoning, synthesis, and modern cheminformatics. With the accelerating shift toward data-driven and AI-assisted molecular science, the ability to rapidly generate Lewis-like representations, including RDKit-parsable SMILES, has become essential for large-scale analysis and predictive modeling. While mature tools exist for organic molecules, transition-metal complexes (TMCs) remain systematically underserved owing to their inherently diverse coordination patterns. To address this gap, we develop TMC-LewisGen, an automated tool for generating Lewis-like structures of TMCs directly from Cartesian coordinates. The task is formulated as a discrete variable optimization problem, where a comprehensive set of explicit chemical rules defines the feasible solution space and integer linear programming (ILP) identifies a globally optimal assignment of bond orders, lone pairs, and formal charges, followed by postprocessing corrections. The final outputs include Lewis-like structures, RDKit-parsable SMILES, and interpretive bonding reports framed within the covalent bond classification (CBC) formalism, in which ligands are categorized as L-, X-, or Z- type donors. Benchmarking on nearly 29,000 TMCs demonstrates robust performance, with valid SMILES generated for 100% of the structures and subsecond ILP solution times in most cases. Representative examples, including Werner-type complexes, η-coordinated π-ligands, and metal–carbene motifs, demonstrate broad applicability. The framework supports large-scale structural annotation, automated chemical interpretation, and data-driven modeling of TMCs. A freely accessible web server for 3D TMC analysis and visualization is available at https://tmc-lewis-structure.streamlit.app/.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"13 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149210145","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":"Hashi: Bridging Statistical Model Derived 1D Microstate Encodings and Protein 3D Structural Ensembles","authors":"Harshavardhan Madhan,Athi N. Naganathan","doi":"10.1021/acs.jcim.6c01760","DOIUrl":"https://doi.org/10.1021/acs.jcim.6c01760","url":null,"abstract":"The functioning of proteins is intimately linked to the conformational states they sample within the native ensemble. Generating ensembles from a single static structure is therefore a research domain receiving considerable attention. In this application note, we introduce Hashi, a pipeline to rapidly generate realistic structural ensembles from the outputs of the structure-based Wako–Saitô–Muñoz–Eaton (WSME) statistical mechanical model of protein folding. This approach relies on integrating the block WSME model outputs─strings of zeros and ones describing the conformational status of every residue over thousands or millions of microstates, each assigned a statistical weight derived from physically grounded energy-entropy terms, and free energy profiles─with the RANCH module of the ensemble optimization method from the ATSAS software suite, providing three-dimensional views of the structural ensembles within the model framework. It is applicable to a variety of single-chain monomeric systems with lengths ranging from 30 to 500 residues, including globular and repeat proteins. Ensembles can be generated within seconds to minutes on a standard computer, without recourse to high-performance computational clusters. The generated structural ensembles can also be rank-ordered according to their free energies within a given macrostate or a range of reaction coordinate values. Since the statistical weights of the WSME model microstates can be reweighted or calibrated with experiments, the ensembles shed light on not just the folding mechanism but also on the structural excursions that determine function and opening of otherwise buried binding pockets.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"167 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149152113","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":"Deciphering Cellular and Spatial Regulatory Identities from Sparse Chromatin Accessibility Landscapes","authors":"Yunhe Wang,Songyu Guo,Jingjing Qian,Heqiu Liang,Yanchi Su,Zhuohan Yu,Xiangtao Li","doi":"10.1021/acs.jcim.6c02880","DOIUrl":"https://doi.org/10.1021/acs.jcim.6c02880","url":null,"abstract":"Chromatin accessibility profiling provides a powerful view of regulatory landscapes across heterogeneous biological systems, enabling the investigation of cell-state diversity and spatially organized regulatory programs. However, deciphering meaningful regulatory identities from sparse and noisy accessibility landscapes remains challenging, particularly for closely related cellular states and neighboring regulatory domains. Unlike transcriptomic measurements, chromatin accessibility does not directly capture canonical marker expression, and regulatory identities must therefore be inferred from indirect evidence, including gene activity scores, transcription factor motif activity, and reference-based annotation. These signals often become ambiguous, limiting the reliability of unsupervised clustering followed by post-hoc interpretation. Here, we present CASSIA (Chromatin Accessibility Semi-Supervised Identity-Aware representation learning), a unified label-guided framework for decoding cellular and spatial regulatory identities from chromatin accessibility landscapes. CASSIA leverages limited high-confidence anchors to learn informative accessibility representations while preserving intrinsic regulatory structures. The framework integrates a binary accessibility denoising autoencoder, a Jensen–Shannon divergence-based clustering constraint, and a similarity difference-aware triplet objective; spatial neighborhood consistency is further incorporated to preserve regulatory organization when spatial information is available. Across diverse chromatin accessibility datasets, CASSIA reconstructs cellular and spatial regulatory identities with improved agreement to biological annotations and tissue organization compared with representative baselines. Identity-associated accessible regions and transcription factor activity profiles reveal distinct regulatory programs underlying inferred identities, demonstrating that CASSIA transforms sparse accessibility landscapes into interpretable cellular and tissue-level regulatory maps.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"12 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149152071","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 Transferable Descriptor for Accurate Prediction of Excitonic Couplings in A-D-A Organic Photovoltaic Acceptors","authors":"Zirui Wang,Qiming Zhao,Yuanping Yi,Lingyun Zhu,Zhixiang Wei","doi":"10.1021/acs.jcim.6c02209","DOIUrl":"https://doi.org/10.1021/acs.jcim.6c02209","url":null,"abstract":"Exciton diffusion length (LD) is a key parameter determining the power conversion efficiency of organic solar cells (OSCs), yet its fast and accurate calculation is computationally challenging due to thin-film structural disorder and the long-range nature of excitonic coupling. Herein, we introduce a coarse-grained Coulomb matrix (CG-CM) descriptor that enables rapid and accurate prediction of excitonic couplings. By mapping molecular conjugated backbones onto a fixed number of beads, our descriptor eliminates the dimensionality dependence on the number of atoms, allowing the machine learning model to generalize across molecules with different sizes and chemical structures. The model achieves high accuracy with a coefficient of determination (R2) exceeding 0.92 across diverse nonfullerene acceptor systems, including Y-type and ITIC-type materials. We further demonstrate the model’s reliability in capturing excitonic coupling fluctuations, predicting LD in disordered thin films, and reproducing LD anisotropy in crystalline systems. This work provides a powerful and efficient tool for high-throughput screening of high-performance organic photovoltaic materials.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"32 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149152070","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":"Latent Style-Based\u0000Quantum Wasserstein GAN for Drug\u0000Design","authors":"Julien Baglio,Yacine Haddad,Richard Polifka","doi":"10.1021/acs.jcim.6c01249","DOIUrl":"https://doi.org/10.1021/acs.jcim.6c01249","url":null,"abstract":"Abstract The development of new drugs is a tedious, time-consuming, and expensive process for which the average costs are estimated to be up to around $2.5 billion. The first step in this long process is the design of the new drug, for which de novo drug design, assisted by artificial intelligence, has blossomed in recent years and has revolutionized the field. In particular, generative artificial intelligence has delivered promising results in drug discovery and development, reducing costs and time to solution. However, classical generative models, such as generative adversarial networks (GANs), can be difficult to train and are prone to mode collapse. Quantum computing may be an avenue to overcome some of these issues, providing compact models with far fewer trainable parameters and thereby enhancing the generalizability of GANs. We propose a new latent style-based quantum Wasserstein GAN (QGAN) for drug design, combining a variational autoencoder that maps molecular structures into a latent space with a hybrid quantum GAN. Our pipeline is benchmarked against the classical models of the industry-relevant MOSES suite. Style-based noise reuploading at every rotational gate and a Wasserstein gradient penalty are used to improve trainability and to mitigate mode collapse. Training and large-scale sampling with circuits of up to 15 qubits are performed on noiseless quantum simulators to validate our architecture. We present an inference-only demonstration on a 156-qubit IBM Heron quantum computer using a five-qubit setup, sampling 2,500 new molecules, to investigate the effect of quantum noise. Our default QGAN, using 110 trainable generator parameters, reaches statistically comparable overall performance and statistically significant gains on selected drug-relevant metrics relative to both the MOSES classical latent GAN baseline, which contains about 700,000 trainable parameters for the generator, and a classical latent GAN with matched generator capacity.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"32 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149210148","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":"RNADaemon: Reinforcement Learning-Guided Molecular Dynamics for RNA Tertiary Structure Prediction.","authors":"Wentao Xu, You Wu, Xiang Yu","doi":"10.1021/acs.jcim.6c02247","DOIUrl":"https://doi.org/10.1021/acs.jcim.6c02247","url":null,"abstract":"<p><p>Accurate prediction of RNA tertiary structure remains a central challenge in structural biology, yet existing computational approaches face fundamental limitations arising from the ruggedness of the RNA free-energy landscape and the scarcity of experimentally resolved structures. Here we present RNADaemon, a framework that formulates RNA tertiary structure prediction as a Markov Decision Process and trains a Soft Actor-Critic agent to apply adaptive external force-field biases that steer all-atom molecular dynamics toward native-like conformations to minimize the eRMSD metric. Trained on 479 monomeric RNA structures and evaluated on 14 test cases from CASP16, RNA-Puzzles, and the PDB, the agent trains stably, approaches SimRNA and FARFAR2 in median RMSD, and recovers the global fold of most reference structures. However, performance remains markedly inferior to that of AlphaFold 3 and trRosettaRNA. We identify the scalar Bellman error as a fundamental information bottleneck that currently limits the performance of the agent. To our knowledge, this is the first application of continuous-action deep reinforcement learning (RL) with all-atom force fields to RNA tertiary structure prediction. This study establishes a benchmark for future comparisons and provides diagnostic insights for RL-guided RNA structure modeling.</p>","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":" ","pages":""},"PeriodicalIF":6.4,"publicationDate":"2026-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149177910","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":"Hiccup: A Resource-Efficient Automated Platform for High-Performance Global Neural Network Potentials Training","authors":"Cheng Chen,Dingming Chen,Haifeng Wang","doi":"10.1021/acs.jcim.6c02301","DOIUrl":"https://doi.org/10.1021/acs.jcim.6c02301","url":null,"abstract":"Neural network potentials (NNPs) enable atomistic simulations with near-first-principles accuracy at greatly reduced cost, yet the construction of specialized high-accuracy NNPs for complex systems remains limited by expert-dependent workflows and the challenge of achieving high accuracy and robust generalization under limited computational budgets. Here we present Hiccup, a resource-efficient automated platform for training high-performance global NNPs. Hiccup leverages prior guidance from universal NNPs to overcome the cold-start bottleneck of active learning, and integrates this initialization with hierarchical genetic algorithm–molecular dynamics (GA–MD) sampling to enable both broad configurational exploration and targeted refinement of low-energy basins. Building on these algorithmic advances, the platform further implements an asynchronous submit-and-forget scheduling mechanism that minimizes resource idling by overlapping first-principles labeling with subsequent sampling, training, and evaluation on the GPU. We demonstrate the application and efficiency of Hiccup using the Cu–Y–O ternary system as a case study. The resulting system-specific NNP achieves near-first-principles accuracy in static predictions and exhibits robust long-time dynamical fidelity across a broad thermodynamic range. These results establish Hiccup as a practical and scalable infrastructure for automated construction of high-quality interatomic potentials for chemically complex systems.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"1 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149152074","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":"Drug-Likeness, Druggability, and Molecular Developability via Chemometric Evaluation: A Review","authors":"Beibei Yu,Haoyu Zhu,Rongling Gu,Shijie Du,Wei Shi,Lu Xu","doi":"10.1021/acs.jcim.6c02860","DOIUrl":"https://doi.org/10.1021/acs.jcim.6c02860","url":null,"abstract":"High attrition rates constitute a major bottleneck in small-molecule drug discovery, largely arising from poor molecular developability rather than an insufficient pharmacological activity. Drug-likeness, druggability, and molecular developability represent complementary evaluation concepts that address different aspects of small-molecule drug discovery. Nevertheless, inconsistent terminology, fragmented evaluation frameworks, and inherent drawbacks of traditional empirical workflows impair the reliability and applicability of molecular property assessment. As powerful multivariate analytical tools, chemometric methods overcome these limitations and enable quantitative, systematic, and high-throughput molecular property evaluation. This Review clarifies the definitions and relationships among the three concepts and summarizes their classical evaluation approaches. Following the paradigm shift from empirical rules to chemometric modeling, multiobjective optimization, and large-model prediction, advanced computational pipelines are categorized into direct molecular similarity evaluation and predictive modeling of key molecular properties. We discuss the principles, progress, performance, strengths, and limitations of each technical strategy and highlight their synergistic values in lead discovery, structural modification, multiobjective optimization, and focused library design. Finally, challenges for standardized molecular-evaluation systems are analyzed, and future trends of chemometrics-assisted multiobjective optimization and large-model-driven intelligent molecular profiling are prospected.","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"64 1","pages":""},"PeriodicalIF":5.6,"publicationDate":"2026-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149152072","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":"Diff-SelfA-PepGSI: A Diffusion-Based Framework for De Novo Generation, Screening, and Morphological Identification of Self-Assembled Peptide Sequences","authors":"Weizhong Sun,&nbsp;Yuning Ma,&nbsp;Xiumin Shi,&nbsp;Mingshan Wei,&nbsp;Ao He,&nbsp;Weizhi Wang","doi":"10.1021/acs.jcim.6c02911","DOIUrl":"10.1021/acs.jcim.6c02911","url":null,"abstract":"<p>Peptide self-assembly is a ubiquitous molecular phenomenonin whichshort amino acid sequences spontaneously organize into diverse nanostructures.These assemblies exhibit distinct morphologies closely linked to theirbiological functions and potential applications, making them attractivecandidates for biomaterials in drug delivery, regenerative medicine,and diagnostics. However, designing and predicting peptide sequenceswith specific self-assembled morphologies remains highly challengingdue to the complex interplay of noncovalent interactions, includinghydrogen bonding, hydrophobic effects, and π–πstacking. Although deep learning has enabled accurate prediction ofpeptide self-assembly propensity, rational <em>de novo</em> design of peptides with targeted assembly behavior remains largelyunexplored. In this study, we propose Diff-SelfA-PepGSI, a novel pipelinefor the morphology-specific design of self-assembling peptides. Thisthree-stage framework comprises: (i) generation, in which a diffusion-basedgenerative model produces candidate self-assembling peptide sequences;(ii) screening, in which pretrained protein language models extractdeep semantic features and a deep discriminative network performssequence optimization and property screening; and (iii) identification,in which a dual-branch structure extracts features and a dedicatedclassifier identifies the specific nanostructures formed. By integratingsequence generation, predictive screening, and morphology identification,this workflow establishes a proof-of-concept <em>de novo</em> design pipeline that links peptide sequences directly to their targetmorphologies. Diff-SelfA-PepGSI demonstrates superior performancecompared to state-of-the-art models in both screening and identificationtasks. Experimental validation preliminarily demonstrates the feasibilityof the <em>de novo</em> design pipeline, successfully generatingself-assembling peptides such as LAPFLA. Under the reported experimentalconditions, 4 of 11 generated candidates (36.4%) exhibited the targetmorphology predicted by the model. Within the challenging contextof <em>de novo</em> morphology-specific peptide design, thissuccess rate supports the potential of Diff-SelfA-PepGSI as an earlystage self-assembling peptide generation platform.</p>","PeriodicalId":44,"journal":{"name":"Journal of Chemical Information and Modeling ","volume":"66 19","pages":"12783–12799"},"PeriodicalIF":6.4,"publicationDate":"2026-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148985071","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}