Multi-task bioassay pre-training for protein-ligand binding affinity prediction
Briefings in Bioinformatics · 22.11.2023 · 25 (1)
DOI: 10.1093/bib/bbad451 ↗Kurz gefasst
Fachartikel in der Zeitschrift Briefings in Bioinformatics, online erschienen am 11.12.2023. Verfasst von Jiaxian Yan und 6 weiteren Autor:innen. Beteiligte Einrichtung: Anhui Province Key Lab of Big Data Analysis and Application, University of Science and Technology of China, JinZhai Road, 230026, Anhui, China. Hauptthemen laut MEDLINE: Proteins, Drug Discovery. Die Arbeit wurde bisher 26-mal zitiert (Crossref) und nennt 101 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, 2022 Tencent Rhino-Bird Research Elite Training Program.
Abstract (Originalsprache)
Protein–ligand binding affinity (PLBA) prediction is the fundamental task in drug discovery. Recently, various deep learning-based models predict binding affinity by incorporating the three-dimensional (3D) structure of protein–ligand complexes as input and achieving astounding progress. However, due to the scarcity of high-quality training data, the generalization ability of current models is still limited. Although there is a vast amount of affinity data available in large-scale databases such as ChEMBL, issues such as inconsistent affinity measurement labels (i.e. IC50, Ki, Kd), different experimental conditions, and the lack of available 3D binding structures complicate the development of high-precision affinity prediction models using these data. To address these issues, we (i) propose Multi-task Bioassay Pre-training (MBP), a pre-training framework for structure-based PLBA prediction; (ii) construct a pre-training dataset called ChEMBL-Dock with more than 300k experimentally measured affinity labels and about 2.8M docked 3D structures. By introducing multi-task pre-training to treat the prediction of different affinity labels as different tasks and classifying relative rankings between samples from the same bioassay, MBP learns robust and transferrable structural knowledge from our new ChEMBL-Dock dataset with varied and noisy labels. Experiments substantiate the capability of MBP on the structure-based PLBA prediction task. To the best of our knowledge, MBP is the first affinity pre-training model and shows great potential for future development. MBP web-server is now available for free at: https://huggingface.co/spaces/jiaxianustc/mbp.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Anhui Province Key Lab of Big Data Analysis and Application, University of Science and Technology of China, JinZhai Road, 230026, Anhui, China.
- Tencent Quantum Laboratory, Tencent, Shennan Road, 518057, Guangdong, China.
Förderung
- National Natural Science Foundation of China
- 2022 Tencent Rhino-Bird Research Elite Training Program
Zitieren
APA
Yan, J., Ye, Z., Yang, Z., Lu, C., Zhang, S., Liu, Q., & Qiu, J. (2023). Multi-task bioassay pre-training for protein-ligand binding affinity prediction. Briefings in Bioinformatics, 25(1). https://doi.org/10.1093/bib/bbad451
Vancouver
Yan J, Ye Z, Yang Z, Lu C, Zhang S, Liu Q, et al. Multi-task bioassay pre-training for protein-ligand binding affinity prediction. Briefings in Bioinformatics. 2023;25(1). doi:10.1093/bib/bbad451
BibTeX
@article{Yan2023,
author = {Yan, Jiaxian and Ye, Zhaofeng and Yang, Ziyi and Lu, Chengqiang and Zhang, Shengyu and Liu, Qi and Qiu, Jiezhong},
title = {Multi-task bioassay pre-training for protein-ligand binding affinity prediction},
journal = {Briefings in Bioinformatics},
year = {2023},
volume = {25},
number = {1},
doi = {10.1093/bib/bbad451}
}
EUVIMED