Gated-GPS: enhancing protein–protein interaction site prediction with scalable learning and imbalance-aware optimization
Briefings in Bioinformatics · 2025
DOI: 10.1093/bib/bbaf248 ↗Kurz gefasst
Fachartikel in der Zeitschrift Briefings in Bioinformatics, 2025. Verfasst von Xin Gao und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science and Engineering, University of California San Diego, 9500 Gilman Drive, La Jolla, California 92093, USA. Hauptthemen laut MEDLINE: Machine Learning, Protein Interaction Mapping, Proteins, Computational Biology. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 59 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: Zhejiang Province Vanguard Goose-Leading Initiative, National Natural Science Foundation of China, the Research Grants Council of the Hong Kong Special Administrative Region.
Abstract (Originalsprache)
In protein–protein interaction site (PPIS) prediction, existing machine learning models struggle with small datasets, limiting their predictive accuracy for unseen proteins. Additionally, class imbalance in protein complexes, where binding residues constitute a small fraction of all residues, hinders model performance. To address these challenges, we constructed a training dataset 9$\times $ larger than previous benchmarks by filtering the latest protein–protein complex data, improving diversity and generalization. We propose Gated-GPS, a Graph Transformer model with a novel gating mechanism designed to effectively leverage this expanded dataset. Additionally, we integrate cross-entropy loss with Tversky Loss to adjust sensitivity to positive and negative samples, mitigating class imbalance by emphasizing underrepresented binding residues. Experimental results show that Gated-GPS outperforms state-of-the-art (SOTA) models across four test sets. Notably, on the UBTest dataset, designed to evaluate generalization on unbounded proteins, our method improves MCC and AUPRC by 18.5% and 21.4%, respectively, over the previous SOTA. In a case study of snake venom toxin–protein interactions, our model accurately identified interaction sites, demonstrating its potential for therapeutic design and advancing the understanding of complex protein interactions.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Department of Computer Science and Engineering, University of California San Diego, 9500 Gilman Drive, La Jolla, California 92093, USA.
- Department of Computer Science and Engineering, The Chinese University of Hong Kong, Ma Liu Shui, Shatin, Hong Kong SAR, Hong Kong SAR 000000, China.
- Zhejiang Lab, Kechu ang Avenue, Hangzhou Zhejiang 311121, China.
- Hangzhou Institute of Medicine, Chinese Academy of Sciences, 150 Dongfang Street Xiasha, Hangzhou, Zhejiang, 310000, China.
- Institute of Medical Intelligence and XR, The Chinese University of Hong Kong, Ma Liu Shui, Shatin, Hong Kong SAR, Hong Kong SAR 000000, China.
Förderung
- Zhejiang Province Vanguard Goose-Leading Initiative
- National Natural Science Foundation of China
- the Research Grants Council of the Hong Kong Special Administrative Region
Zitieren
APA
Gao, X., Cao, H., Li, J., Qiu, J., Chen, G., & Heng, P. A. (2025). Gated-GPS: enhancing protein–protein interaction site prediction with scalable learning and imbalance-aware optimization. Briefings in Bioinformatics. https://doi.org/10.1093/bib/bbaf248
Vancouver
Gao X, Cao H, Li J, Qiu J, Chen G, Heng PA. Gated-GPS: enhancing protein–protein interaction site prediction with scalable learning and imbalance-aware optimization. Briefings in Bioinformatics. 2025. doi:10.1093/bib/bbaf248
BibTeX
@article{Gao2025,
author = {Gao, Xin and Cao, Hanqun and Li, Jinpeng and Qiu, Jiezhong and Chen, Guangyong and Heng, Pheng-Ann},
title = {Gated-GPS: enhancing protein–protein interaction site prediction with scalable learning and imbalance-aware optimization},
journal = {Briefings in Bioinformatics},
year = {2025},
doi = {10.1093/bib/bbaf248}
}
EUVIMED