reinforcement learning

Robust low-carbon ship routing under weather uncertainty: a distributional reinforcement learning approach

LumiCharge: Spherical Harmonic Convolutional Networks for Atomic Charge Prediction in Drug Discovery.

Atomic charge is crucial in drug design for analyzing reactive sites and interactions between ligands and targets. While quantum mechanical methods offer high accuracy, they are generally computationally costly. Conversely, empirical approaches, while computationally efficient, frequently suffer from lack of precision and generalizability. Recent a number of machine learning-based models have been developed for atomic charge predictions, but they struggle with accurately representing molecular structures and capturing the chemical environments affecting atomic charges, thus limiting their generalization and accuracy. To overcome these limitations, we propose LumiCharge, a novel atomic charge prediction framework that incorporates high-order spherical harmonics convolutions and explicitly models multibody interactions. In constructing this model, we employ a strategy that integrates both high- and low-order information, enhancing its geometric spatial perception capability, which is currently underexplored in the field. Benchmark evaluations demonstrate that LumiCharge outperforms state-of-the-art (SOTA) models by 30%-60% across diverse data sets. Additionally, in cross-scale experiments, LumiCharge demonstrates exceptional extrapolation capability and robustness across molecules of varying sizes, effectively overcoming the limitations imposed by molecular sizes. On an external halogen-containing test set, LumiCharge achieves an RMSE of 0.055e, meeting practical application requirements. Finally, a case study of virtual screening for the androgen receptor (AR) target further validates its outstanding accuracy compared to the OPLS3e force field and other deep learning (DL)-based baseline models, highlighting its exceptional generalization capacity and practical utility in real-world scenarios.

PepBAN: A Deep Learning Framework with Bilinear Attention and Adversarial Learning for Peptide-Protein Interaction Prediction

Accurate prediction of the peptide-protein interaction (PepPI) is crucial for developing peptide-based therapeutics and vaccines. However, this computational task has traditionally faced significant challenges, such as the scarcity of structure data along with the corresponding label of the binding affinity for bound complexes. To address these challenges, we introduce PepBAN, a deep learning framework for modeling PepPI predictions. PepBAN incorporates two technical advancements: (1) adopting the protein language model ESM-2 to characterize proteins and ESM-2 or a graph-based foundation model for peptides without structure data and (2) leveraging the conditional domain adversarial learning to enhance generalization across a broad range of protein targets, especially when there are limited binding data. At the core of PepBAN is a bilinear attention network (BAN) that effectively learns the pattern of pairwise local interactions, enables the identification of key residues participating in the peptide-protein interactions, and offers an intuitive approach to interpret the underlying mechanisms of PepPIs via analyzing attention weights. Our numerical experiments demonstrated that PepBAN outperformed the previous state-of-the-art models across several well-established benchmark studies. Furthermore, we evaluated PepBAN's applicability in predicting cyclic peptide-protein interactions, a task that poses significant challenges due to the presence of noncanonical amino acids. These nonstandard residues require specialized handling, which most existing sequence-based PepPI prediction models did not adequately address, and we adopt an atom-resolved molecular graph approach to process cyclic peptides. Despite this complexity, PepBAN demonstrated a clear advantage by achieving a superior prediction performance and offering a distinct edge in tackling the emerging chemical space of cyclic peptides, which has great potential for novel therapeutic development. In summary, PepBAN serves as a valuable tool for advancing peptide-based drug and therapeutic development.

SynGFN: learning across chemical space with generative flow-based molecular discovery

ChargeNet: E(3) Equivariant Graph Attention Network for Atomic Charge Prediction

Atomic charge is a fundamental quantum chemical property essential for advancing drug design and discovery. Although quantum mechanics (QM) methods offer the highest level of accuracy, their computational demands scale quadratically with the number of atoms, limiting their practicality for large-scale applications. In light of this, empirical and semiempirical methods have been introduced to improve computational efficiency, albeit often at the expense of accuracy. The advent of artificial intelligence has witnessed a growing application of machine learning (ML) techniques to accelerate atomic charge predictions. However, existing ML models often suffer from low accuracy and limited generalization capabilities. To address these challenges, we introduce an advanced equivariant graph attention neural network specifically engineered to model long-range atomic electrostatic interactions with high precision. This model introduces a sophisticated global graph attention mechanism, enabling it to capture charge contributions across multiple scales. By utilizing a combination of structural symmetry-preserving transformations and multiscale attention, our approach not only preserves the inherent symmetries of molecular structures but also substantially improves the model's accuracy, generalization, and robustness in complex scenarios. Our empirical analyses demonstrate that, compared to leading baseline models, the proposed model improves charge prediction accuracy by over 40% on average across various charge-calculation schemes. Remarkably, the model achieves superior performance on the external RESP (restrained electrostatic potential) test data sets, with a 54.6% improvement over the baseline. Additionally, we evaluated our charge model under the setting of virtual screening, where it outperforms both the OPLS3 charges and baseline deep learning models across all evaluation metrics, highlighting its extensive potential for scientific discovery.

PepPCBench is a Comprehensive Benchmarking Framework for Protein-Peptide Complex Structure Prediction

Accurate modeling of protein-peptide interactions is essential for understanding fundamental biological processes and designing peptide-based drugs. However, predicting the complex structures of these interactions remains challenging, primarily due to the high conformational flexibility of peptides. To support a fair and systematic evaluation of recent deep learning (DL) approaches, we introduce PepPCBench, a benchmarking framework tailored to assess protein folding neural networks (PFNNs) in protein-peptide complex prediction. As part of this framework, we curated PepPCSet, a data set of 261 experimentally resolved complexes with peptides ranging from 5 to 30 residues. We benchmark five full-atom PFNNs, including AlphaFold3 (AF3), AlphaFold-Multimer (AFM), Chai-1, HelixFold3 (HF3), and RoseTTAFold-All-Atom (RFAA), using comprehensive evaluation metrics. Our benchmarking reveals meaningful performance differences among these methods and highlights the influence of peptide length, conformational flexibility, and training set similarity on prediction accuracy. While AF3 shows strong performance in structure prediction, further analysis indicates that confidence metrics correlate poorly with experimental binding affinities, underscoring the need for improved scoring strategies and generalizability. By providing a reproducible and extensible framework, PepPCBench enables a robust evaluation of PFNN-based methods and supports their continued development for peptide-protein structure prediction.

CarsiDock-Cov: A deep learning-guided approach for automated covalent docking and screening

The interest in covalent drugs has resurged in recent decades, spurring the development of numerous specialized computational docking tools to facilitate covalent ligand design and screening. Herein, we present CarsiDock-Cov, a new paradigm distinguishing itself as the first deep learning (DL)-guided approach for covalent docking. CarsiDock-Cov retains the core components of its non-covalent predecessor, leveraging a DL model pretrained on millions of docking complexes to predict protein–ligand distance matrices, along with a dedicated-designed geometric optimization procedure to convert these distances into refined binding poses. Additionally, it incorporates several key enhancements specifically tailored to optimize the protocol for covalent docking applications. Our approach has been extensively validated on multiple public datasets regarding the docking and screening of covalent ligands, and the results indicate that our approach not only achieves comparably improved applicability compared to its non-covalent predecessor, but also exhibits competitive performance against various state-of-the-art covalent docking tools. Collectively, our approach represents a significant advance in covalent docking methodology, offering an automated and efficient solution that shows considerable promise for accelerating covalent drug discovery and design.

EpiMII: Integrating Structure and Graph Neural Networks for MHC-II Epitope and Neoantigen Design

MHC-II neoantigens play a critical role in immunotherapy, either as direct effectors or through their influence on CD8+ T cells. However, only a small fraction of tumor DNA mutations qualify as functional neoantigens, and current prediction tools often lack accuracy, leading to the low immunogenicity of predicted neoantigens in vivo. Here, we present EpiMII, a Graph Neural Network model for MHC-II epitope design, which learns from the structural features of epitopes to predict their sequences. To train EpiMII, we constructed a reliable, large dataset containing 142,934 MHC-II epitope structures. This approach achieves a 4.2x improvement over ProteinMPNN, with a sequence recovery rate of 78.0% for known MHC-II epitopes in the Protein Data Bank. As a case study, we designed a neoantigen from hepatocellular carcinoma. All five designed epitopes significantly activated CD4+ T cells in vitro and induced secretion of IFN-γ and TNF-α. Notably, one epitope treatment significantly reduced tumor volume in mice in vivo. EpiMII offers a novel and efficient approach for identifying MHC-II epitopes/neoantigens, potentially contributing to vaccine development.

Q‐GEM: Quantum Chemistry Knowledge Fusion Geometry‐Enhanced Molecular Representation for Property Prediction

Recently, various self‐supervised learning (SSL) methods based on 3D graph neural networks (GNNs) have been developed to comprehensively represent the structural information of molecules in 3D space; this is essential for discovering new drugs. However, existing methods fail to comprehensively characterize the 3D structures of molecules and neglect the electronic structural information that significantly influences key properties such as molecular reactivity, strong electrostatic interactions, and chemical adsorption. Therefore, here, a novel molecular representation learning method is constructed, Q‐GEM, incorporating quantum and geometric structural information enhancement, based on the quantum chemical property database QuanDB and SSL methods. Q‐GEM comprises a GNN embedded with the molecular electronic and complete 3D geometrical structural information as well as several well‐designed multiscale SSL tasks, achieving superior absolute molecular conformation prediction and conformational discrimination. The Q‐GEM achieved state‐of‐the‐art performance in 12 out of 13 prediction tasks on the MoleculeNet dataset, with an average performance improvement of 3.3% and 2.0% for classification and regression prediction tasks, respectively. Moreover, an average performance improvement of 5.2% is achieved in three localized quantum chemical properties, fully demonstrating the excellent performance of Q‐GEM in distinguishing molecular electronic structures. The Q‐GEM represents a novel, powerful breakthrough for accurate molecular property prediction.

Effective generation of heavy-atom-free triplet photosensitizers containing multiple intersystem crossing mechanisms based on deep learning

Photodynamic therapy (PDT) is a clinically approved therapeutic modality that has demonstrated significant potential for cancer treatment, and triplet photosensitizers (PSs) play a key role in its efficacy. Despite deep learning having emerged as a next-generation tool for material discovery, existing methods mainly target a limited subset of triplet PSs, such as thermally activated delayed fluorescence (TADF) materials, neglecting the critical intersystem crossing (ISC) between the high-lying singlet and triplet states (ΔESnTn). To overcome this limitation, we compiled a comprehensive dataset (∼1.90 × 109) of triplet PSs encompassing various ISC mechanisms. Then, we proposed a novel strategy that incorporates two models: a fragment-based model (Frag-MD) and a character-based model (MD), both integrating a conditional transformer, recurrent neural networks, and reinforcement learning. In silico experiments revealed that the Frag-MD model outperforms the MD model in generating larger conjugated motifs with higher average ring numbers and atom counts; while the MD model generates twice as many unique motifs and excels in novelty and diversity, as evaluated by conditional and MOSES metrics. Therefore, our approach is highly effective for modifying conjugated motifs and designing novel triplet PSs. Notably, the recently reported high-efficiency triplet PSs have been re-identified through ablation experiments using our proposed models, which target ΔESnTn and significantly outperform traditional baselines, achieving a prediction accuracy of 73% versus 4%. Our approach holds the potential to establish a new paradigm for discovering novel PSs applicable in PDT.

Revisiting Protein-Protein Docking: A Systematic Evaluation Framework

Protein-protein interactions play pivotal roles in a wide range of biological processes. Determining the atomic-level structures of protein-protein complexes is indispensable for elucidating macromolecular interaction mechanisms and advancing structure-based drug design. Protein-protein docking, as one of the leading computational approaches for predicting complex structures, has seen considerable progress but requires rigorous evaluation in practical applications. In this study, we proposed a comprehensive benchmarking framework to evaluate 11 docking methods spanning traditional (HDOCK, PatchDock, PIPER, ZDOCK) and deep learning (DL)-based (EquiDock, ElliDock, EBMDock, GeoDock, DiffDock-PP, AlphaFold-Multimer, AlphaFold3) approaches. Our framework incorporates the classical DockingBenchmark 5.5 data set for evaluating flexible docking, introduces a newly curated data set (AACBench) for antibody-antigen complex docking, and establishes the PPCBench data set to examine the out-of-distribution (OOD) generalization capabilities of DL-based methods. In docking against apo structures, AlphaFold3 achieves a superior top-5 success rate of 77.98%, whereas the traditional approach HDOCK reaches merely 12.84%, despite its highest top-5 success rate of 85.24% when docking against holo structures. For antibody-antigen docking, AlphaFold3 remains the most accurate method (top-5 success rate: 31.78%) and substantially outperforms AlphaFold-Multimer in modeling the CDR-H3 loop. In OOD generalization tests, all DL-based models exhibit markedly reduced performance on the PPCBench data set. Overall, our work establishes a unified benchmarking framework that enables systematic evaluation of docking methods across diverse tasks and provides critical insights into the strengths and limitations of current docking strategies, thereby informing future developments in protein-protein docking research.

MetalloDock: Decoding Metalloprotein-Ligand Interactions via Physics-Aware Deep Learning for Metalloprotein Drug Discovery.

Accurate prediction of metalloprotein-ligand interactions is critical for metalloprotein-targeted drug discovery. Conventional docking tools and existing deep learning (DL) models fail to reliably capture metal-ligand interactions, hampering the discovery of potent metalloprotein inhibitors. Here, we propose MetalloDock, the first DL-based docking framework specially designed for metalloprotein targets. By innovatively integrating an autoregressive spatial decoding engine with a physics-constrained geometric generation paradigm, MetalloDock can precisely reconstruct metal coordination geometries and accurately capture metal-ligand interactions, which enhance both the accuracy of metalloprotein-ligand docking and binding affinity prediction. Extensive evaluations on our custom-built benchmark data set demonstrate that MetalloDock outperforms existing methods, including AlphaFold3, in docking success rate and virtual screening performance for metalloprotein targets. In real-world applications, MetalloDock successfully identified multiple novel hit compounds in a virtual screening campaign targeting the prostate-specific membrane antigen. Additionally, it enabled rational drug design for acidic polymerase endonuclease, leading to the discovery of potent inhibitors. These results highlight the broad applicability of MetalloDock in accelerating metalloprotein-targeted drug discovery and provide a standardized framework for future evaluation of metalloprotein-specific docking algorithms.