Design of New STAT3 Inhibitors Based on RNN Algorithms and Molecular Simulations
10.13865/j.cnki.cjbmb.2024.09.1183
- VernacularTitle:基于RNN算法和分子模拟设计新的STAT3抑制剂
- Author:
Yi-Kuan XUE
1
;
Yu-Xiang WANG
1
;
Pei WANG
1
;
Wei-Zhong LU
1
Author Information
1. 苏州科技大学 电子与信息工程学院,江苏 苏州215009
- Publication Type:Journal Article
- Keywords:
inhibitor;
signal transducer and activator of transcription3 (STAT3);
recurrent neural net-work(RNN)
- From:
Chinese Journal of Biochemistry and Molecular Biology
2024;40(12):1732-1741
- CountryChina
- Language:Chinese
-
Abstract:
Signal transducer and activator of transcription 3 (STAT3) is aberrantly expressed in a variety of cancer stem cells,and it strongly correlated with cellular carcinogenesis.Therefore,designing and in-vestigating new STAT3 inhibitors is an important and effective strategy to combat cancer.In this paper,we propose a novel STAT3 inhibitor design method based on the recurrent neural network (RNN) algo-rithm,and evaluate the effectiveness of this method through molecular simulation studies.We first used the RNN algorithm to construct a STAT3 inhibitor generation model,so that it can generate brand new in-hibitors.Then based on machine learning algorithms,we established a molecular classification prediction model for STAT3 inhibitors,conducted a hierarchical virtual screening based on molecular docking on molecules classified as STAT3 inhibitors,and selected three molecules with the highest scores in the extra precision (XP) screen,which as potential inhibitors for the next step of the study.The potential inhibi-tors were subjected to binding free energy computation,and tprediction of absorption,distribution,me-tabolism,excretion toxicity (ADMET) .In addition,an independent gradient model (IGM) analysis was used to further explore the pharmacogenetic properties.The experimental results in this paper show that novel potential STAT3 inhibitors with good drugability can be effectively generated by using the above methods.In sum,our work can provide a reference for subsequent STAT3 drug development.