Graph neural networks (GNNs) have been widely used in molecular property prediction, but explaining their black-box predictions is still a challenge. Most existing explanation methods for GNNs in chemistry focus on attributing model predictions to individual nodes, edges or fragments that are not necessarily derived from a chemically meaningful segmentation of molecules. To address this challenge, we propose a method named substructure mask explanation (SME). SME is based on well-established molecular segmentation methods and provides an interpretation that aligns with the understanding of chemists. We apply SME to elucidate how GNNs learn to predict aqueous solubility, genotoxicity, cardiotoxicity and blood–brain barrier permeation for small molecules. SME provides interpretation that is consistent with the understanding of chemists, alerts them to unreliable performance, and guides them in structural optimization for target properties. Hence, we believe that SME empowers chemists to confidently mine structure-activity relationship (SAR) from reliable GNNs through a transparent inspection on how GNNs pick up useful signals when learning from data.

Large language models (LLMs) have shown remarkable advancements in chemistry and biomedical research, acting as versatile foundation models for various tasks. We introduce AMP-Designer, an LLM-based approach, for swiftly designing antimicrobial peptides (AMPs) with desired properties. Within 11 days, AMP-Designer achieved the de novo design of 18 AMPs with broad-spectrum activity against Gram-negative bacteria. In vitro validation revealed a 94.4% success rate, with two candidates demonstrating exceptional antibacterial efficacy, minimal hemotoxicity, stability in human plasma, and low potential to induce resistance, as evidenced by significant bacterial load reduction in murine lung infection experiments. The entire process, from design to validation, concluded in 48 days. AMP-Designer excels in creating AMPs targeting specific strains despite limited data availability, with a top candidate displaying a minimum inhibitory concentration of 2.0 micrograms per milliliter against Propionibacterium acnes . Integrating advanced machine learning techniques, AMP-Designer demonstrates remarkable efficiency, paving the way for innovative solutions to antibiotic resistance.

Applying machine learning algorithms to protein–ligand scoring functions has aroused widespread attention in recent years due to the high predictive accuracy and affordable computational cost. Nevertheless, most machine learning-based scoring functions are only applicable to a specific task, e.g., binding affinity prediction, binding pose prediction or virtual screening, suggesting that the development of a scoring function with balanced performance in all critical tasks remains a grand challenge. To this end, we propose a novel parameterization strategy by introducing an adjustable binding affinity term that represents the correlation between the predicted outcomes and experimental data into the training of mixture density network. The resulting residue-atom distance likelihood potential not only retains the superior docking and screening power over all the other state-of-the-art approaches, but also achieves a remarkable improvement in scoring and ranking performance. We emphatically explore the impacts of several key elements on prediction accuracy as well as the task preference, and demonstrate that the performance of scoring/ranking and docking/screening tasks of a certain model could be well balanced through an appropriate manner. Overall, our study highlights the potential utility of our innovative parameterization strategy as well as the resulting scoring framework in future structure-based drug design.

The expertise accumulated in deep neural network-based structure prediction has been widely transferred to the field of protein–ligand binding pose prediction, thus leading to the emergence of a variety of deep learning-guided docking models for predicting protein–ligand binding poses without relying on heavy sampling. However, their prediction accuracy and applicability are still far from satisfactory, partially due to the lack of protein–ligand binding complex data. To this end, we create a large-scale complex dataset containing ∼9 M protein–ligand docking complexes for pre-training, and propose CarsiDock, the first deep learning-guided docking approach that leverages pre-training of millions of predicted protein–ligand complexes. CarsiDock contains two main stages, i.e., a deep learning model for the prediction of protein–ligand atomic distance matrices, and a translation, rotation and torsion-guided geometry optimization procedure to reconstruct the matrices into a credible binding pose. The pre-training and multiple innovative architectural designs facilitate the dramatically improved docking accuracy of our approach over the baselines in terms of multiple docking scenarios, thereby contributing to its outstanding early recognition performance in several retrospective virtual screening campaigns. Further explorations demonstrate that CarsiDock can not only guarantee the topological reliability of the binding poses but also successfully reproduce the crucial interactions in crystalized structures, highlighting its superior applicability.