Residential College | false |
Status | 已發表Published |
Chemical toxicity prediction based on semi-supervised learning and graph convolutional neural network | |
Chen, Jiarui1; Si, Yain Whar1; Un, Chon Wai1; Siu, Shirley W.I.1,2,3 | |
2021-12-01 | |
Source Publication | Journal of Cheminformatics |
ISSN | 1758-2946 |
Volume | 13Issue:1 |
Abstract | As safety is one of the most important properties of drugs, chemical toxicology prediction has received increasing attentions in the drug discovery research. Traditionally, researchers rely on in vitro and in vivo experiments to test the toxicity of chemical compounds. However, not only are these experiments time consuming and costly, but experiments that involve animal testing are increasingly subject to ethical concerns. While traditional machine learning (ML) methods have been used in the field with some success, the limited availability of annotated toxicity data is the major hurdle for further improving model performance. Inspired by the success of semi-supervised learning (SSL) algorithms, we propose a Graph Convolution Neural Network (GCN) to predict chemical toxicity and trained the network by the Mean Teacher (MT) SSL algorithm. Using the Tox21 data, our optimal SSL-GCN models for predicting the twelve toxicological endpoints achieve an average ROC-AUC score of 0.757 in the test set, which is a 6% improvement over GCN models trained by supervised learning and conventional ML methods. Our SSL-GCN models also exhibit superior performance when compared to models constructed using the built-in DeepChem ML methods. This study demonstrates that SSL can increase the prediction power of models by learning from unannotated data. The optimal unannotated to annotated data ratio ranges between 1:1 and 4:1. This study demonstrates the success of SSL in chemical toxicity prediction; the same technique is expected to be beneficial to other chemical property prediction tasks by utilizing existing large chemical databases. Our optimal model SSL-GCN is hosted on an online server accessible through: https://app.cbbio.online/ssl-gcn/home. |
Keyword | Admet Chemical Toxicity Deep Learning Graph Convolutional Neural Network Mean Teacher Semi-supervised Learning Tox21 |
DOI | 10.1186/s13321-021-00570-8 |
URL | View the original |
Indexed By | SCIE |
Language | 英語English |
WOS Research Area | Chemistry ; Computer Science |
WOS Subject | Chemistry, Multidisciplinary ; Computer Science, Information Systems ; Computer Science, Interdisciplinary Applications |
WOS ID | WOS:000722997100001 |
Scopus ID | 2-s2.0-85119986794 |
Fulltext Access | |
Citation statistics | |
Document Type | Journal article |
Collection | DEPARTMENT OF COMPUTER AND INFORMATION SCIENCE |
Corresponding Author | Siu, Shirley W.I. |
Affiliation | 1.Department of Computer and Information Science, University of Macau, Taipa, Avenida da Universidade, 999078, Macao 2.Institute of Science and Environment, University of Saint Joseph, Rua de Londres 106, 999078, Macao 3.School of Pharmaceutical Sciences, Universiti Sains Malaysia, Penang, USM, 11800, Malaysia |
First Author Affilication | University of Macau |
Corresponding Author Affilication | University of Macau |
Recommended Citation GB/T 7714 | Chen, Jiarui,Si, Yain Whar,Un, Chon Wai,et al. Chemical toxicity prediction based on semi-supervised learning and graph convolutional neural network[J]. Journal of Cheminformatics, 2021, 13(1). |
APA | Chen, Jiarui., Si, Yain Whar., Un, Chon Wai., & Siu, Shirley W.I. (2021). Chemical toxicity prediction based on semi-supervised learning and graph convolutional neural network. Journal of Cheminformatics, 13(1). |
MLA | Chen, Jiarui,et al."Chemical toxicity prediction based on semi-supervised learning and graph convolutional neural network".Journal of Cheminformatics 13.1(2021). |
Files in This Item: | There are no files associated with this item. |
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.
Edit Comment