基于人工智能和互联网时代的化合物毒性预测  

Toxicity Prediction Based on Artificial Intelligence and the Internet Era

作　　者：徐涛 秦越 单伟 贾辰阳 朱加良 邓志光 李卓玥 刘丹会 Xu Tao;Qin Yue;ShanWei;Jia Chenyang;Zhu Jialiang;Deng Zhiguang;Li Zhuoyue;Liu Danhui(Science and Technology on Reactor System Design Technology Laboratory,Nuclear Power Institute of China,Chengdu,610213;Key Laboratory of Pesticide&Chemical Biology,Ministry of Education,College of Chemistry,Central China Normal University,Wuhan,430079)

机构地区：[1]中国核动力研究设计院核反应堆系统设计技术重点实验室,成都610213 [2]华中师范大学化学学院农药与化学生物学教育部重点实验室,武汉430079

出　　处：《化学通报》2023年第3期312-322,共11页Chemistry

摘　　要：随着商品中所含各种化合物的不断使用,人们日益关注其对人类及生态环境的安全危害。在过去的几年里,通过计算方法预测化合物毒性已经显示出极大的潜力。在此,总结了常用的机器学习和深度学习算法在建立毒性预测模型上的优缺点,并系统回顾了近三年发表的可免费访问的毒性预测网络服务器。此外,还讨论了基于人工智能和互联网时代下毒性预测所面临的机遇和挑战。希望指导人们合理选择算法和网络服务器进行建模及化合物毒性评估。With the continuous using compounds contained in commodities,there is growing concern about their harm to human and ecological environment safety.In the past few years,computational techniques have showed their potential to predict toxicity of compounds.Here,we summarize the advantages and drawbacks of machine learning and deep learning algorithms for establishing toxicity prediction models,and systematically review the freely accessible toxicity prediction web servers for in silico toxicity prediction in the past three years.Additionally,the opportunities and challenges of toxicity prediction based on artificial intelligence and internet are discussed.It is hoped that this paper can provide help in guiding people to rationally choose algorithms and web servers for modeling and toxicity evaluation.

关 键 词：人工智能 深度学习 机器学习 毒性预测 网络服务器 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP393.09[自动化与计算机技术—控制科学与工程] R99[医药卫生—毒理学]