Department of Pharmaceutical Chemistry, Department of Bioengineering & Therapeutic Sciences, Institute for Neurodegenerative Diseases, Kavli Institute for Fundamental Neuroscience, Bakar Computational Health Sciences Institute, University of California, San Francisco, 675 Nelson Rising Lane, San Francisco, California 94143, United States.
J Med Chem. 2020 Aug 27;63(16):8705-8722. doi: 10.1021/acs.jmedchem.0c00385. Epub 2020 May 15.
The accurate modeling and prediction of small molecule properties and bioactivities depend on the critical choice of molecular representation. Decades of informatics-driven research have relied on expert-designed molecular descriptors to establish quantitative structure-activity and structure-property relationships for drug discovery. Now, advances in deep learning make it possible to efficiently and compactly molecular representations directly from data. In this review, we discuss how active research in molecular deep learning can address limitations of current descriptors and fingerprints while creating new opportunities in cheminformatics and virtual screening. We provide a concise overview of the role of representations in cheminformatics, key concepts in deep learning, and argue that learning representations provides a way forward to improve the predictive modeling of small molecule bioactivities and properties.
小分子性质和生物活性的准确建模和预测取决于分子表示的关键选择。几十年来，基于信息学的研究依赖于专家设计的分子描述符来建立药物发现的定量结构-活性和结构-性质关系。现在，深度学习的进步使得直接从数据中高效紧凑地表示分子成为可能。在这篇综述中，我们讨论了分子深度学习的活跃研究如何解决当前描述符和指纹的局限性，同时在化学信息学和虚拟筛选中创造新的机会。我们简要概述了表示法在化学信息学中的作用、深度学习的关键概念，并认为学习表示法为提高小分子生物活性和性质的预测建模提供了一种方法。
