The arc of drug discovery entails a multiparameter optimization problem spanning vast length scales. The key parameters range from solubility (angstroms) to protein–ligand binding (nanometers) to in vivo toxicity (meters). Through feature learning—instead of feature engineering—deep neural networks promise to outperform both traditional physics-based and knowledge-based machine learning models for predicting molecular properties pertinent to drug discovery. To this end, we present the PotentialNet family of graph convolutions. These models are specifically designed for and achieve state-of-the-art performance for protein–ligand binding affinity. We further validate these deep neural networks by setting new standards of performance in several ligand-based tasks. In parallel, we introduce a new metric, the Regression Enrichment Factor EF_χ^(R), to measure the early enrichment of computational models for chemical data. Finally, we introduce a cross-validation strategy based on structural homology clustering that can more accurately measure model generalizability, which crucially distinguishes the aims of machine learning for drug discovery from standard machine learning tasks.

Defense systems are usually deployed to protect high-value targets or hot spots that are integral parts of the modern battlefield environment. However, in coastal defense operations (due to the variability of the maritime environment and the sustainability of combat), limited operational capabilities, the need for efficient coordination, and protracted combat are peculiarly challenging to meet by traditional manned fleets. In contrast, with lower costs, unmanned fleets can organize an autonomous defense against enemy targets that are capable of rapid response. This paper focuses on the typical defense scenario; we analyzed and modeled the objective functions of the intelligent defense system and propose a hierarchical distributed multi-agent-based system design scheme. Finally, to test the system’s performance, we established simulation verification experiments in a typical scenario and compared the system based on the traditional central architecture. The results show that, in a defense operation, the hierarchically-distributed multi-agent-based system shows improvements in system decision-making efficiency and interception effect.

Alzheimer’s disease (AD) is a chronic and progressive neurodegenerative disorder and the pathogenesis of AD is poorly understood. G protein-coupled receptors (GPCRs) are involved in numerous key AD pathways and play a key role in the pathology of AD. To fully understand the pathogenesis of AD and design novel drug therapeutics, analyzing the connection between AD and GPCRs is of great importance. In this paper, we firstly build and analyze the AD-related pathway by consulting the KEGG pathway of AD and a mass of literature and collect 25 AD-related GPCRs for drug discovery. Then the ILbind and AutoDock Vina tools are integrated to find out potential drugs related to AD. According to the analysis of DUD-E dataset, we select five drugs, that is, Acarbose (ACR), Carvedilol (CVD), Digoxin (DGX), NADH (NAI), and Telmisartan (TLS), by sorting the ILbind scores (≥0.73). Then depending on their AutoDock Vina scores and pocket position information, the binding patterns of these five drugs are obtained. We analyze the regulation function of GPCRs in the metabolic network of AD based on the drug screen results, which may be helpful for the study of the off-target effect and the side effect of drugs.

Many ligands simultaneously interact with multiple protein chains in quaternary structure (QS). However, a significant number of previous studies on template-based modeling of protein-ligand interactions were based on monomeric structure (MS), which may suffer from incomplete binding information. The defects of using MS rather than QS have not been systematically studied before. In this work, based on molecular docking experiments and binding free energy estimations, we performed a large-scale comparison of the protein-ligand interactions in both forms of structures. We found that 1) about 18.6% biologically relevant ligands bind multiple chains in QS simultaneously. 2) For more than 95% complexes with multiple chains involved in the interactions, the binding free energy is lower for the QS form than the MS form. 3) For over 70% complexes with multi-chain binding pockets, docking with QS yields more accurate ligand conformations than with MS. While for about 1.82% complexes, accurate docking conformations were obtained by MS. Based on this work, it is encouraged to make use of QS rather than MS in future studies on protein-ligand interactions.