人工智能药物发现研究体系的构建及实践  

Construction and practice of artificial intelligence in drug discovery system

作　　者：张志豪 周吟玦 姜慧君[1] 蒋南[1] ZHANG Zhi-hao;ZHOU Yin-jue;JIANG Hui-jun;JIANG Nan(School of Pharmacy,Nanjing Medical University,Nanjing 211166,China)

机构地区：[1]南京医科大学药学院,南京211166

出　　处：《中国新药杂志》2022年第13期1294-1303,共10页Chinese Journal of New Drugs

基　　金：江苏省自然科学基金资助项目(BK20211254);南京医科大学教育研究课题(2019LX020)。

摘　　要：新型靶向药物的研发具有费用高、周期长、成功率低的特点,其最大瓶颈在于研发过程中存在诸多不确定性因素,如靶点有效性、模型的可靠性等问题,需要通过大量实验予以确认。而在药物研发过程中引入人工智能技术,应用机器学习及深度学习算法提取分子结构特征、分析药物-靶标相互作用、构建药物-疾病-蛋白质之间的联系,可以在不同研发环节建立具有较高准确率的预测系统,并可减少各个研发环节的不确定性,从而缩短研发周期、降低试错成本、提高研发成功率。本文综述了机器学习和深度学习在人工智能药物研发中的构建及实践,以期为新药研发提供理论依据。The research and development(R&D)of new-type targeted drugs are characteristic of high cost,long cycle with low success rate.The most important bottleneck is that there are many uncertain factors in the R&D process,such as effectiveness of target,reliability of model and so on,which need to be confirmed by a large number of experiments.With the introduction of artificial intelligence into the process of drug R&D,machine learning and deep learning algorithms can be used to extract the features of molecular structures,analyze drug-target interactions,and construct drug-disease-protein relationships.These are helpful to establish the prediction systems with high accuracy and reduce the uncertainty of each stage in R&D,so as to shorten the R&D cycle,reduce trial-and-error costs and improve the success rate.This article summarizes the application of artificial intelligence in drug discovery system constructed by machine learning and deep learning methods in the field of drug R&D,hoping to provide a theoretical basis for new drug development.

关 键 词：机器学习 深度学习 人工智能 新药发现 

分 类 号：R939[医药卫生—生药学] R918[医药卫生—药学]