ZeroBind: a protein-specific zero-shot predictor with subgraph matching for drug-target interactions
Abstract Existing drug-target interaction (DTI) prediction methods generally fail to generalize well to novel (unseen) proteins and drugs. In this study, we propose a protein-specific meta-learning framework ZeroBind with subgraph matching for predicting protein-drug interactions from their structur...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Nature Portfolio
2023-11-01
|
Series: | Nature Communications |
Online Access: | https://doi.org/10.1038/s41467-023-43597-1 |
_version_ | 1797414700350177280 |
---|---|
author | Yuxuan Wang Ying Xia Junchi Yan Ye Yuan Hong-Bin Shen Xiaoyong Pan |
author_facet | Yuxuan Wang Ying Xia Junchi Yan Ye Yuan Hong-Bin Shen Xiaoyong Pan |
author_sort | Yuxuan Wang |
collection | DOAJ |
description | Abstract Existing drug-target interaction (DTI) prediction methods generally fail to generalize well to novel (unseen) proteins and drugs. In this study, we propose a protein-specific meta-learning framework ZeroBind with subgraph matching for predicting protein-drug interactions from their structures. During the meta-training process, ZeroBind formulates training a protein-specific model, which is also considered a learning task, and each task uses graph neural networks (GNNs) to learn the protein graph embedding and the molecular graph embedding. Inspired by the fact that molecules bind to a binding pocket in proteins instead of the whole protein, ZeroBind introduces a weakly supervised subgraph information bottleneck (SIB) module to recognize the maximally informative and compressive subgraphs in protein graphs as potential binding pockets. In addition, ZeroBind trains the models of individual proteins as multiple tasks, whose importance is automatically learned with a task adaptive self-attention module to make final predictions. The results show that ZeroBind achieves superior performance on DTI prediction over existing methods, especially for those unseen proteins and drugs, and performs well after fine-tuning for those proteins or drugs with a few known binding partners. |
first_indexed | 2024-03-09T05:37:21Z |
format | Article |
id | doaj.art-03b89d87e3a54f44abb798ab3eba76b8 |
institution | Directory Open Access Journal |
issn | 2041-1723 |
language | English |
last_indexed | 2024-03-09T05:37:21Z |
publishDate | 2023-11-01 |
publisher | Nature Portfolio |
record_format | Article |
series | Nature Communications |
spelling | doaj.art-03b89d87e3a54f44abb798ab3eba76b82023-12-03T12:27:37ZengNature PortfolioNature Communications2041-17232023-11-0114111410.1038/s41467-023-43597-1ZeroBind: a protein-specific zero-shot predictor with subgraph matching for drug-target interactionsYuxuan Wang0Ying Xia1Junchi Yan2Ye Yuan3Hong-Bin Shen4Xiaoyong Pan5Institute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University, and Key Laboratory of System Control and Information Processing, Ministry of Education of ChinaInstitute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University, and Key Laboratory of System Control and Information Processing, Ministry of Education of ChinaDepartment of Computer Science and Engineering, and MoE Key Lab of Artificial Intelligence, AI Institute, Shanghai Jiao Tong UniversityInstitute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University, and Key Laboratory of System Control and Information Processing, Ministry of Education of ChinaInstitute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University, and Key Laboratory of System Control and Information Processing, Ministry of Education of ChinaInstitute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University, and Key Laboratory of System Control and Information Processing, Ministry of Education of ChinaAbstract Existing drug-target interaction (DTI) prediction methods generally fail to generalize well to novel (unseen) proteins and drugs. In this study, we propose a protein-specific meta-learning framework ZeroBind with subgraph matching for predicting protein-drug interactions from their structures. During the meta-training process, ZeroBind formulates training a protein-specific model, which is also considered a learning task, and each task uses graph neural networks (GNNs) to learn the protein graph embedding and the molecular graph embedding. Inspired by the fact that molecules bind to a binding pocket in proteins instead of the whole protein, ZeroBind introduces a weakly supervised subgraph information bottleneck (SIB) module to recognize the maximally informative and compressive subgraphs in protein graphs as potential binding pockets. In addition, ZeroBind trains the models of individual proteins as multiple tasks, whose importance is automatically learned with a task adaptive self-attention module to make final predictions. The results show that ZeroBind achieves superior performance on DTI prediction over existing methods, especially for those unseen proteins and drugs, and performs well after fine-tuning for those proteins or drugs with a few known binding partners.https://doi.org/10.1038/s41467-023-43597-1 |
spellingShingle | Yuxuan Wang Ying Xia Junchi Yan Ye Yuan Hong-Bin Shen Xiaoyong Pan ZeroBind: a protein-specific zero-shot predictor with subgraph matching for drug-target interactions Nature Communications |
title | ZeroBind: a protein-specific zero-shot predictor with subgraph matching for drug-target interactions |
title_full | ZeroBind: a protein-specific zero-shot predictor with subgraph matching for drug-target interactions |
title_fullStr | ZeroBind: a protein-specific zero-shot predictor with subgraph matching for drug-target interactions |
title_full_unstemmed | ZeroBind: a protein-specific zero-shot predictor with subgraph matching for drug-target interactions |
title_short | ZeroBind: a protein-specific zero-shot predictor with subgraph matching for drug-target interactions |
title_sort | zerobind a protein specific zero shot predictor with subgraph matching for drug target interactions |
url | https://doi.org/10.1038/s41467-023-43597-1 |
work_keys_str_mv | AT yuxuanwang zerobindaproteinspecificzeroshotpredictorwithsubgraphmatchingfordrugtargetinteractions AT yingxia zerobindaproteinspecificzeroshotpredictorwithsubgraphmatchingfordrugtargetinteractions AT junchiyan zerobindaproteinspecificzeroshotpredictorwithsubgraphmatchingfordrugtargetinteractions AT yeyuan zerobindaproteinspecificzeroshotpredictorwithsubgraphmatchingfordrugtargetinteractions AT hongbinshen zerobindaproteinspecificzeroshotpredictorwithsubgraphmatchingfordrugtargetinteractions AT xiaoyongpan zerobindaproteinspecificzeroshotpredictorwithsubgraphmatchingfordrugtargetinteractions |