人工智能在宿主与病原体蛋白质互作预测中的应用进展  

Advances in the Application of Artificial Intelligence to Predict Host-Pathogen Protein Interactions

作　　者：任碧燕 刘璐[1] 舒坤贤[1] 曾垂省[1] 代劲[2] 刘川[1] 李娜[2] REN Biyan;LIU Lu;SHU Kunxian;ZENG Chuisheng;DAI Jin;LIU Chuan;LI Na(Chongqing Key Laboratory of Big Data for Bio Intelligence,School of Life Health Information Science and Engineering,Chongqing University of Posts and Telecommunications,Chongqing 400065,China;School of Software Engineering,Chongqing University of Posts and Telecommunications,Chongqing400065,China)

机构地区：[1]重庆邮电大学生命健康信息科学与工程学院大数据生物智能重庆市重点实验室,重庆400065 [2]重庆邮电大学软件工程学院,重庆400065

出　　处：《病毒学报》2024年第5期1121-1136,共16页Chinese Journal of Virology

基　　金：重庆市教委科学技术研究项目(项目号:KJQN202300616),题目:面向宿主与病原菌互作的机器学习关键技术研究。

摘　　要：宿主与病原体之间的蛋白质-蛋白质互作(Host-pathogen protein-protein interaction,HP-PPI)是病原体感染宿主的关键分子事件,准确识别HP-PPI对于理解宿主的免疫防御机制、病原体的致病机制,以及研发抗感染药物都具有重要意义。近年来,蛋白质互作实验技术的发展及其在宿主与病原体互作研究中的应用,积累了大量的HPPPI数据,于是人工智能技术逐渐在HP-PPI预测研究领域中脱颖而出。本文综述了人工智能方法在HP-PPI预测研究领域中的应用,首先介绍了基于人工智能方法识别HP-PPI的任务流程,总结了收录HP-PPI数据的常用数据库;然后重点对机器学习和深度学习两大类人工智能方法在HP-PPI预测研究领域中的应用进行分类归纳,介绍了部分经典算法模型的基本原理、特征选择和模型评估结果等;最后,分析了人工智能方法预测HP-PPI面临的问题及挑战,以期为宿主与病原体互作研究领域的科研人员提供思路和参考。Host-pathogen protein-protein interaction(HP-PPI)is a key molecular event in host during infection by pathogens.Elucidation of HP-PPI is crucial for understanding the immune defense mechanism of the host,the mechanism of pathogenesis,and development of anti-infection drugs.In recent years,the development of PPI detection methods and their application in host-pathogen interaction studies have accumulated a large amount of HP-PPI data,which help the artificial intelligence(AI)emerge as outstanding techniques in the research field of HP-PPI prediction.This paper reviews the application of AI techniques in HP-PPI prediction.Firstly,the workflow of AI-aided identification of HP-PPI is outlined,and the commonly used databases containing HP-PPI data are summarized.Subsequently,we focus on the application of two major categories of AI methods,namely machine learning and deep learning,in the research field of HP-PPI prediction,and present the fundamentals of several classical algorithmic models,feature selection methods and model evaluation results.Finally,the problems and challenges faced by AI-aided HP-PPI prediction were discussed in detail to provide insights for researchers in studying host-pathogen interactions.

关 键 词：宿主-病原体互作 蛋白质-蛋白质互作 机器学习 深度学习 

分 类 号：Q-1[生物学]