Learning the Interaction Prior for Protein-Protein Interaction Prediction: A Model-Agnostic Approach

Ziqi Gao, Chenyi Zi, Zijing Liu, Ziqiao Meng, Yu Li, Jia Li
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:34015-34029, 2026.

Abstract

Protein-protein interactions (PPIs) are fundamental to cellular function, disease mechanisms, and drug discovery. Current learning-based PPI predictors focus on learning powerful protein representations but neglect designing specialized classification heads. They mainly rely on generic aggregating methods like concatenation or dot products, which lack biological insight. Motivated by the biological "L3 rule", where multiple length-3 paths between a pair of proteins indicate their interaction likelihood, our study addresses this gap by designing a biologically informed PPI classifier. In this paper, we provide empirical evidence that popular PPI datasets strongly support the L3 rule. We propose an L3-path-regularized graph prompt learning method called L3-PPI, which can generate a prompt graph with virtual L3 paths based on protein representations and controls the number of paths. L3-PPI reformulates the classification of protein embedding pairs into a graph-level classification task over the generated prompt graph. This lightweight module seamlessly integrates with leading PPI predictors as a plug-and-play component, injecting the interaction prior of complementarity to enhance performance. Extensive experiments show that L3-PPI achieves superior performance enhancements over state-of-the-art competitors.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-gao26ap, title = {Learning the Interaction Prior for Protein-Protein Interaction Prediction: A Model-Agnostic Approach}, author = {Gao, Ziqi and Zi, Chenyi and Liu, Zijing and Meng, Ziqiao and Li, Yu and Li, Jia}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {34015--34029}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/gao26ap/gao26ap.pdf}, url = {https://proceedings.mlr.press/v306/gao26ap.html}, abstract = {Protein-protein interactions (PPIs) are fundamental to cellular function, disease mechanisms, and drug discovery. Current learning-based PPI predictors focus on learning powerful protein representations but neglect designing specialized classification heads. They mainly rely on generic aggregating methods like concatenation or dot products, which lack biological insight. Motivated by the biological "L3 rule", where multiple length-3 paths between a pair of proteins indicate their interaction likelihood, our study addresses this gap by designing a biologically informed PPI classifier. In this paper, we provide empirical evidence that popular PPI datasets strongly support the L3 rule. We propose an L3-path-regularized graph prompt learning method called L3-PPI, which can generate a prompt graph with virtual L3 paths based on protein representations and controls the number of paths. L3-PPI reformulates the classification of protein embedding pairs into a graph-level classification task over the generated prompt graph. This lightweight module seamlessly integrates with leading PPI predictors as a plug-and-play component, injecting the interaction prior of complementarity to enhance performance. Extensive experiments show that L3-PPI achieves superior performance enhancements over state-of-the-art competitors.} }
Endnote
%0 Conference Paper %T Learning the Interaction Prior for Protein-Protein Interaction Prediction: A Model-Agnostic Approach %A Ziqi Gao %A Chenyi Zi %A Zijing Liu %A Ziqiao Meng %A Yu Li %A Jia Li %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-gao26ap %I PMLR %P 34015--34029 %U https://proceedings.mlr.press/v306/gao26ap.html %V 306 %X Protein-protein interactions (PPIs) are fundamental to cellular function, disease mechanisms, and drug discovery. Current learning-based PPI predictors focus on learning powerful protein representations but neglect designing specialized classification heads. They mainly rely on generic aggregating methods like concatenation or dot products, which lack biological insight. Motivated by the biological "L3 rule", where multiple length-3 paths between a pair of proteins indicate their interaction likelihood, our study addresses this gap by designing a biologically informed PPI classifier. In this paper, we provide empirical evidence that popular PPI datasets strongly support the L3 rule. We propose an L3-path-regularized graph prompt learning method called L3-PPI, which can generate a prompt graph with virtual L3 paths based on protein representations and controls the number of paths. L3-PPI reformulates the classification of protein embedding pairs into a graph-level classification task over the generated prompt graph. This lightweight module seamlessly integrates with leading PPI predictors as a plug-and-play component, injecting the interaction prior of complementarity to enhance performance. Extensive experiments show that L3-PPI achieves superior performance enhancements over state-of-the-art competitors.
APA
Gao, Z., Zi, C., Liu, Z., Meng, Z., Li, Y. & Li, J.. (2026). Learning the Interaction Prior for Protein-Protein Interaction Prediction: A Model-Agnostic Approach. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:34015-34029 Available from https://proceedings.mlr.press/v306/gao26ap.html.

Related Material