{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,29]],"date-time":"2026-05-29T14:41:03Z","timestamp":1780065663965,"version":"3.54.0"},"reference-count":60,"publisher":"Oxford University Press (OUP)","issue":"4","license":[{"start":{"date-parts":[[2025,8,28]],"date-time":"2025-08-28T00:00:00Z","timestamp":1756339200000},"content-version":"vor","delay-in-days":58,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["31871274"],"award-info":[{"award-number":["31871274"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Natural Science Foundation of Chongqing, China","award":["CSTB2022NSCQ-MSX0650"],"award-info":[{"award-number":["CSTB2022NSCQ-MSX0650"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,7,2]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>The prediction of binary protein\u2013protein interactions (PPIs) is essential for protein engineering, but a major challenge in deep learning-based methods is the unknown decision-making process of the model. To address this challenge, we propose the ESM2_AMP framework, which utilizes the ESM2 protein language model for extracting segment features from actual amino acid sequences and integrates the Transformer model for feature fusion in binary PPIs prediction. Further, the two distinct models, ESM2_AMPS and ESM2_AMP_CSE are developed to systematically explore the contributions of segment features and combine with special tokens features in the decision-making process. The experimental results reveal that the model relying on segment features demonstrates strong correlations between segments with high attention weights and known functional regions of amino acid sequences. This insight suggests that attention to these segments helps capture biologically relevant functional and interaction-related information. By analyzing the coverage relationship between high-attention sequence fragments and functional regions, we validated the model\u2019s ability to capture key segment features of PPIs and revealed the critical role of functional domains in PPIs. This finding not only enhances the interpretability methods for sequence-based prediction models but also provides biological evidence supporting the important regulatory role of functional sequences in protein\u2013protein interactions. It offers cross-disciplinary insights for algorithm optimization and experimental validation research in the field of computational biology.<\/jats:p>","DOI":"10.1093\/bib\/bbaf434","type":"journal-article","created":{"date-parts":[[2025,8,28]],"date-time":"2025-08-28T12:38:47Z","timestamp":1756384727000},"source":"Crossref","is-referenced-by-count":8,"title":["ESM2_AMP: an interpretable framework for protein\u2013protein interactions prediction and biological mechanism discovery"],"prefix":"10.1093","volume":"26","author":[{"given":"Yawen","family":"Sun","sequence":"first","affiliation":[{"name":"College of Life Science, Chongqing Normal University , No. 37 University Town Road, high-tech District, Chongqing 401331 ,","place":["P.R. China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rui","family":"Wang","sequence":"additional","affiliation":[{"name":"College of Life Science, Chongqing Normal University , No. 37 University Town Road, high-tech District, Chongqing 401331 ,","place":["P.R. China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6650-9975","authenticated-orcid":false,"given":"Zeyu","family":"Luo","sequence":"additional","affiliation":[{"name":"College of Life Science, Chongqing Normal University , No. 37 University Town Road, high-tech District, Chongqing 401331 ,","place":["P.R. China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lejia","family":"Tan","sequence":"additional","affiliation":[{"name":"College of Life Science, Chongqing Normal University , No. 37 University Town Road, high-tech District, Chongqing 401331 ,","place":["P.R. China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Junhao","family":"Liu","sequence":"additional","affiliation":[{"name":"College of Life Science, Chongqing Normal University , No. 37 University Town Road, high-tech District, Chongqing 401331 ,","place":["P.R. China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ruimeng","family":"Li","sequence":"additional","affiliation":[{"name":"College of Life Science, Chongqing Normal University , No. 37 University Town Road, high-tech District, Chongqing 401331 ,","place":["P.R. China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4200-7502","authenticated-orcid":false,"given":"Dongqing","family":"Wei","sequence":"additional","affiliation":[{"name":"State Key Laboratory of Microbial Metabolism, Joint International Research Laboratory of Metabolic & Developmental Sciences and School of Life Sciences and Biotechnology, Shanghai Jiao Tong University , 800 Dongchuan RD. Minhang District, Shanghai 200030 ,","place":["P.R. China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6361-0840","authenticated-orcid":false,"given":"Yu-Juan","family":"Zhang","sequence":"additional","affiliation":[{"name":"College of Life Science, Chongqing Normal University , No. 37 University Town Road, high-tech District, Chongqing 401331 ,","place":["P.R. China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"286","published-online":{"date-parts":[[2025,8,28]]},"reference":[{"key":"2025082808384079900_ref1","doi-asserted-by":"publisher","first-page":"6501","DOI":"10.1016\/j.cell.2024.10.038","article-title":"Discovery and significance of protein-protein interactions in health and disease","volume":"187","author":"Greenblatt","year":"2024","journal-title":"Cell"},{"key":"2025082808384079900_ref2","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbad076","article-title":"Machine learning on protein-protein interaction prediction: models, challenges and trends","volume":"24","author":"Tang","year":"2023","journal-title":"Brief Bioinform"},{"key":"2025082808384079900_ref3","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbab036","article-title":"A survey on computational models for predicting protein-protein interactions","volume":"22","author":"Hu","year":"2021","journal-title":"Brief Bioinform"},{"key":"2025082808384079900_ref4","doi-asserted-by":"publisher","first-page":"1105","DOI":"10.1007\/s00424-014-1672-2","article-title":"The role of protein\u2013protein interactions in the intracellular traffic of the potassium channels TASK-1 and TASK-3","volume":"467","author":"Kilisch","year":"2015","journal-title":"Pfl\u00fcgers Archiv - European Journal of Physiology"},{"key":"2025082808384079900_ref5","doi-asserted-by":"publisher","first-page":"27(1)","DOI":"10.3390\/molecules27010041","article-title":"Benchmark evaluation of protein-protein interaction prediction algorithms","author":"Dunham","year":"2021","journal-title":"Molecules"},{"key":"2025082808384079900_ref6","doi-asserted-by":"publisher","first-page":"245","DOI":"10.1038\/340245a0","article-title":"Song O-k, a novel genetic system to detect protein\u2013protein interactions","volume":"340","author":"Fields","year":"1989","journal-title":"Nature"},{"key":"2025082808384079900_ref7","doi-asserted-by":"publisher","first-page":"211","DOI":"10.1007\/978-1-0716-3445-5_18","volume-title":"Springer","author":"Lin","year":"2017"},{"key":"2025082808384079900_ref8","doi-asserted-by":"publisher","first-page":"3025","DOI":"10.1093\/nar\/gkn159","article-title":"Using support vector machine combined with auto covariance to predict protein-protein interactions from protein sequences","volume":"36","author":"Guo","year":"2008","journal-title":"Nucleic Acids Res"},{"key":"2025082808384079900_ref9","doi-asserted-by":"publisher","first-page":"4337","DOI":"10.1073\/pnas.0607879104","article-title":"Predicting protein-protein interactions based only on sequences information","volume":"104","author":"Shen","year":"2007","journal-title":"Proc Natl Acad Sci U S A"},{"key":"2025082808384079900_ref10","doi-asserted-by":"publisher","first-page":"i802","DOI":"10.1093\/bioinformatics\/bty573","article-title":"Predicting protein-protein interactions through sequence-based deep learning","volume":"34","author":"Hashemifar","year":"2018","journal-title":"Bioinformatics"},{"key":"2025082808384079900_ref11","doi-asserted-by":"publisher","first-page":"i305","DOI":"10.1093\/bioinformatics\/btz328","article-title":"Multifaceted protein-protein interaction prediction based on Siamese residual RCNN","volume":"35","author":"Chen","year":"2019","journal-title":"Bioinformatics"},{"key":"2025082808384079900_ref12","doi-asserted-by":"publisher","first-page":"e7126","DOI":"10.7717\/peerj.7126","article-title":"An integration of deep learning with feature embedding for protein-protein interaction prediction","volume":"7","author":"Yao","year":"2019","journal-title":"PeerJ"},{"key":"2025082808384079900_ref13","doi-asserted-by":"publisher","first-page":"7112","DOI":"10.1109\/TPAMI.2021.3095381","article-title":"ProtTrans: toward understanding the language of life through self-supervised learning","volume":"44","author":"Elnaggar","year":"2022","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"2025082808384079900_ref14","doi-asserted-by":"publisher","first-page":"106","DOI":"10.1186\/s12859-024-05725-6","article-title":"xCAPT5: protein-protein interaction prediction using deep and wide multi-kernel pooling convolutional neural networks with protein language model","volume":"25","author":"Dang","year":"2024","journal-title":"BMC Bioinformatics"},{"key":"2025082808384079900_ref15","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbae359","article-title":"TUnA: an uncertainty-aware transformer model for sequence-based protein-protein interaction prediction","volume":"25","author":"Ko","year":"2024","journal-title":"Brief Bioinform"},{"key":"2025082808384079900_ref16","doi-asserted-by":"publisher","first-page":"1093","DOI":"10.1038\/s41467-023-36736-1","article-title":"Hierarchical graph learning for protein\u2013protein interaction","volume":"14","author":"Gao","year":"2023","journal-title":"Nat Commun"},{"key":"2025082808384079900_ref17","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1706.03762","article-title":"Attention is all you need","author":"Vaswani","year":"2017"},{"key":"2025082808384079900_ref18","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1810.04805","article-title":"BERT: pre-training of deep bidirectional transformers for language understanding","author":"Devlin","year":"2018"},{"key":"2025082808384079900_ref19","doi-asserted-by":"publisher","first-page":"1123","DOI":"10.1126\/science.ade2574","article-title":"Evolutionary-scale prediction of atomic-level protein structure with a language model","volume":"379","author":"Lin","year":"2023","journal-title":"Science"},{"key":"2025082808384079900_ref20","doi-asserted-by":"publisher","first-page":"885","DOI":"10.1038\/s41586-024-08370-4","article-title":"Site-saturation mutagenesis of 500 human protein domains","volume":"637","author":"Beltran","year":"2025","journal-title":"Nature"},{"key":"2025082808384079900_ref21","doi-asserted-by":"publisher","first-page":"1427","DOI":"10.1038\/s42256-023-00756-9","article-title":"Learning characteristics of graph neural networks predicting protein\u2013ligand affinities","volume":"5","author":"Mastropietro","year":"2023","journal-title":"Nature Machine Intelligence"},{"key":"2025082808384079900_ref22","doi-asserted-by":"publisher","first-page":"126327","DOI":"10.1016\/j.neucom.2023.126327","article-title":"Deep neural networks in the cloud: review, applications, challenges and research directions","volume":"545","author":"Chan","year":"2023","journal-title":"Neurocomputing"},{"key":"2025082808384079900_ref23","doi-asserted-by":"publisher","first-page":"4992","DOI":"10.1021\/pr100618t","article-title":"Large-scale prediction of human protein\u2212protein interactions from amino acid sequence based on latent topic features","volume":"9","author":"Pan","year":"2010","journal-title":"J Proteome Res"},{"key":"2025082808384079900_ref24","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbae076","article-title":"Cracking the black box of deep sequence-based protein-protein interaction prediction","volume":"25","author":"Bernett","year":"2024","journal-title":"Brief Bioinform"},{"key":"2025082808384079900_ref25","doi-asserted-by":"publisher","first-page":"1454","DOI":"10.1038\/s41592-024-02359-7","article-title":"Applying interpretable machine learning in computational biology\u2014pitfalls, recommendations and opportunities for new developments","volume":"21","author":"Chen","year":"2024","journal-title":"Nat Methods"},{"key":"2025082808384079900_ref26","first-page":"4768","volume-title":"A Unified Approach to Interpreting Model Predictions, in Proceedings of the 31st International Conference on Neural Information Processing Systems","author":"Lundberg","year":"2017"},{"key":"2025082808384079900_ref27","doi-asserted-by":"publisher","first-page":"652","DOI":"10.1038\/s42003-022-03617-0","article-title":"Protein-protein interaction and non-interaction predictions using gene sequence natural vector","volume":"5","author":"Zhao","year":"2022","journal-title":"Commun Biol"},{"key":"2025082808384079900_ref28","doi-asserted-by":"publisher","first-page":"D396","DOI":"10.1093\/nar\/gkt1079","article-title":"Negatome 2.0: a database of non-interacting proteins derived by literature mining, manual annotation and protein structure analysis","volume":"42","author":"Blohm","year":"2014","journal-title":"Nucleic Acids Res"},{"key":"2025082808384079900_ref29","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1311.1714","volume-title":"KaHIP v3.00 -- Karlsruhe High Quality Partitioning -- User Guide","author":"Sanders","year":"2013"},{"key":"2025082808384079900_ref30","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbad534","article-title":"Interpretable feature extraction and dimensionality reduction in ESM2 for protein localization prediction","volume":"25","author":"Luo","year":"2024","journal-title":"Brief Bioinform"},{"key":"2025082808384079900_ref31","doi-asserted-by":"publisher","first-page":"91","DOI":"10.1016\/B978-0-12-821379-7.00003-5","volume-title":"Practical Machine Learning for Data Analysis Using Python","author":"Subasi","year":"2020"},{"key":"2025082808384079900_ref32","doi-asserted-by":"publisher","first-page":"1115","DOI":"10.24272\/j.issn.2095-8137.2023.263","article-title":"From beasts to bytes: revolutionizing zoological research with artificial intelligence","volume":"44","author":"Zhang","year":"2023","journal-title":"Zool Res"},{"key":"2025082808384079900_ref33","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1711.05101","article-title":"Fixing weight decay regularization in Adam","author":"Loshchilov","year":"2017"},{"key":"2025082808384079900_ref34","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1907.10902","volume-title":"Optuna: A Next-Generation Hyperparameter Optimization Framework","author":"Akiba","year":"2019"},{"key":"2025082808384079900_ref35","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2310.01758","volume-title":"Linearization of ReLU Activation Function for Neural Network-Embedded Optimization:Optimal Day-Ahead Energy Scheduling","author":"Zhao","year":"2023"},{"key":"2025082808384079900_ref36","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1502.01852","volume-title":"Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification","author":"He","year":"2015"},{"key":"2025082808384079900_ref37","doi-asserted-by":"publisher","first-page":"4","DOI":"10.1186\/s13040-023-00322-4","article-title":"The Matthews correlation coefficient (MCC) should replace the ROC AUC as the standard metric for assessing binary classification","volume":"16","author":"Chicco","year":"2023","journal-title":"BioData Min"},{"key":"2025082808384079900_ref38","first-page":"3","volume-title":"Autoencoders, Minimum Description Length and Helmholtz Free Energy, in Proceedings of the 7th International Conference on Neural Information Processing Systems","author":"Hinton","year":"1993"},{"key":"2025082808384079900_ref39","doi-asserted-by":"publisher","first-page":"56","DOI":"10.1038\/s42256-019-0138-9","article-title":"From local explanations to global understanding with explainable AI for trees","volume":"2","author":"Lundberg","year":"2020","journal-title":"Nat Mach Intell"},{"key":"2025082808384079900_ref40","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2003.05991","volume-title":"Autoencoders","author":"Bank D","year":"2020"},{"key":"2025082808384079900_ref41","doi-asserted-by":"publisher","first-page":"31","DOI":"10.17849\/insm-47-01-31-39.1","article-title":"Random Forest","volume":"47","author":"Rigatti","year":"2017","journal-title":"J Insur Med"},{"key":"2025082808384079900_ref42","doi-asserted-by":"publisher","first-page":"369","DOI":"10.1093\/bib\/bbr016","article-title":"Letter to the editor: on the stability and ranking of predictors from random forest variable importance measures","volume":"12","author":"Nicodemus","year":"2011","journal-title":"Brief Bioinform"},{"key":"2025082808384079900_ref43","doi-asserted-by":"publisher","DOI":"10.3390\/biom14030259","article-title":"The configuration of GRB2 in protein interaction and signal transduction","volume":"14","author":"Wang","year":"2024","journal-title":"Biomolecules"},{"key":"2025082808384079900_ref44","doi-asserted-by":"publisher","first-page":"458","DOI":"10.1093\/bib\/bbz007","article-title":"Disentangling the complexity of low complexity proteins","volume":"21","author":"Mier","year":"2020","journal-title":"Brief Bioinform"},{"key":"2025082808384079900_ref45","doi-asserted-by":"publisher","first-page":"27","DOI":"10.3390\/molecules27020423","article-title":"Sequence and structure-based analyses of human Ankyrin repeats","author":"Chakrabarty","year":"2022","journal-title":"Molecules"},{"key":"2025082808384079900_ref46","doi-asserted-by":"publisher","first-page":"26","DOI":"10.1016\/j.sbi.2018.09.009","article-title":"Affinity and specificity of motif-based protein-protein interactions","volume":"54","author":"Ivarsson","year":"2019","journal-title":"Curr Opin Struct Biol"},{"key":"2025082808384079900_ref47","doi-asserted-by":"publisher","first-page":"119654","DOI":"10.1016\/j.bbamcr.2023.119654","article-title":"Disordered regions mediate the interaction of p53 and MRE11","volume":"1871","author":"Usluer","year":"2024","journal-title":"Biochim Biophys Acta Mol Cell Res"},{"key":"2025082808384079900_ref48","doi-asserted-by":"publisher","first-page":"D523","DOI":"10.1093\/nar\/gkac1052","article-title":"UniProt: the universal protein knowledgebase in 2023","volume":"51","author":"UniProt","year":"2023","journal-title":"Nucleic Acids Res"},{"key":"2025082808384079900_ref50","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1412.3555","volume-title":"Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling","author":"Chung","year":"2014"},{"key":"2025082808384079900_ref51","doi-asserted-by":"publisher","first-page":"969","DOI":"10.1016\/j.cels.2021.08.010","article-title":"D-SCRIPT translates genome to phenome with sequence-based, structure-aware, genome-scale predictions of protein-protein interactions","volume":"12","author":"Sledzieski","year":"2021","journal-title":"Cell Syst"},{"key":"2025082808384079900_ref52","doi-asserted-by":"publisher","first-page":"i264","DOI":"10.1093\/bioinformatics\/btac258","article-title":"Topsy-Turvy: integrating a global view into sequence-based PPI prediction","volume":"38","author":"Singh","year":"2022","journal-title":"Bioinformatics"},{"key":"2025082808384079900_ref53","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/W18-5431","volume-title":"An Analysis of Encoder Representations in Transformer-Based Machine Translation. in BlackboxNLP@EMNLP","author":"Raganato","year":"2018"},{"key":"2025082808384079900_ref54","doi-asserted-by":"publisher","first-page":"vbad001","DOI":"10.1093\/bioadv\/vbad001","article-title":"Applications of transformer-based language models in bioinformatics: a survey","volume":"3","author":"Zhang","year":"2023","journal-title":"Bioinform Adv"},{"key":"2025082808384079900_ref55","doi-asserted-by":"publisher","first-page":"e1009736","DOI":"10.1371\/journal.pcbi.1009736","article-title":"Positional SHAP (PoSHAP) for interpretation of machine learning models trained from biological sequences","volume":"18","author":"Dickinson","year":"2022","journal-title":"PLoS Comput Biol"},{"key":"2025082808384079900_ref56","doi-asserted-by":"publisher","DOI":"10.1109\/PAIS62114.2024.10541203","volume-title":"in 2024 6th International Conference on Pattern Analysis and Intelligent Systems (PAIS)","author":"Dehimi","year":"2024"},{"key":"2025082808384079900_ref57","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btad401","article-title":"Generating interacting protein sequences using domain-to-domain translation","volume":"39","author":"Meynard-Piganeau","year":"2023","journal-title":"Bioinformatics"},{"key":"2025082808384079900_ref58","doi-asserted-by":"publisher","DOI":"10.1101\/2024.12.13.628435","volume-title":"Paying Attention to Attention: High Attention Sites as Indicators of Protein Family and Function in Language Models","author":"Nayar","year":"2024"},{"key":"2025082808384079900_ref59","doi-asserted-by":"publisher","first-page":"117","DOI":"10.1006\/jsbi.2001.4392","article-title":"Protein repeats: structures, functions, and evolution","volume":"134","author":"Andrade","year":"2001","journal-title":"J Struct Biol"},{"key":"2025082808384079900_ref60","doi-asserted-by":"publisher","first-page":"D444","DOI":"10.1093\/nar\/gkae1082","article-title":"InterPro: the protein sequence classification resource in 2025","volume":"53","author":"Blum","year":"2025","journal-title":"Nucleic Acids Res"},{"key":"2025082808384079900_ref61","doi-asserted-by":"publisher","DOI":"10.1101\/2024.11.05.622169","article-title":"PLM-interact: extending protein language models to predict protein-protein interactions","author":"Liu","year":"2024","journal-title":"bioRxiv"}],"container-title":["Briefings in Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bib\/article-pdf\/26\/4\/bbaf434\/64148085\/bbaf434.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bib\/article-pdf\/26\/4\/bbaf434\/64148085\/bbaf434.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,8,28]],"date-time":"2025-08-28T12:38:49Z","timestamp":1756384729000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bib\/article\/doi\/10.1093\/bib\/bbaf434\/8242608"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,7]]},"references-count":60,"journal-issue":{"issue":"4","published-print":{"date-parts":[[2025,7,2]]}},"URL":"https:\/\/doi.org\/10.1093\/bib\/bbaf434","relation":{},"ISSN":["1467-5463","1477-4054"],"issn-type":[{"value":"1467-5463","type":"print"},{"value":"1477-4054","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2025,7]]},"published":{"date-parts":[[2025,7]]},"article-number":"bbaf434"}}