生命科学仪器2023，Vol.21Issue(3)：23-35,13.DOI:10.11967/2023210604
深度学习在药物靶点亲和力预测中的应用
Application of Deep Learning in Prediction of Drug Target Affinity in Drug Development
胡少飞 1辛念 2田德振 2李博3
作者信息
- 1. 北京理工大学生命学院,北京 100081||北京理工大学生命学院,生物医药成分分离与分析北京市重点实验室,北京 100081
- 2. 北京理工亘舒科技有限公司,北京 100081
- 3. 北京理工大学生命学院,生物医药成分分离与分析北京市重点实验室,北京 100081||北京理工大学前沿交叉科学研究院,北京,100081
- 折叠
摘要
Abstract
In the early stage of drug research and development,computer methods for predicting the binding affinity of small molecule drugs and protein targets have proved their key role.Such methods are called Drug Target Affinity(DTA)prediction.Among them,the DTA prediction method based on deep learning shows its excellent perform-ance and great potential.This review comprehensively reviews topics related to deep learning based DTA prediction,such as source databases,data mining and feature learning methods,deep learning models,and some representative prediction methods developed using these resources.In this review,we first discussed the data of compounds and proteins from various libraries from the perspective of data formats and coding schemes.For DTA prediction models,we classify them from two perspectives,namely,the perspective of prediction task type and the perspective of learning method adopted by the model,and summarize the representative DTA prediction models of each type.Finally,we discussed some remaining problems to explore and develop more powerful and accurate DTA prediction methods.关键词
药物研发/深度学习/药物-靶点亲和力预测Key words
drug discovery/deep learning/DTA prediction分类
医药卫生引用本文复制引用
胡少飞,辛念,田德振,李博..深度学习在药物靶点亲和力预测中的应用[J].生命科学仪器,2023,21(3):23-35,13.
