{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,14]],"date-time":"2026-07-14T14:10:03Z","timestamp":1784038203500,"version":"3.55.0"},"reference-count":46,"publisher":"Oxford University Press (OUP)","issue":"7","license":[{"start":{"date-parts":[[2023,6,27]],"date-time":"2023-06-27T00:00:00Z","timestamp":1687824000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100012166","name":"National Key R&D Program of China","doi-asserted-by":"publisher","award":["2022YFA1303700"],"award-info":[{"award-number":["2022YFA1303700"]}],"id":[{"id":"10.13039\/501100012166","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["22033001"],"award-info":[{"award-number":["22033001"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["32270689"],"award-info":[{"award-number":["32270689"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2023,7,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:sec>\n                  <jats:title>Motivation<\/jats:title>\n                  <jats:p>In recent years, high-throughput sequencing technologies have made large-scale protein sequences accessible. However, their functional annotations usually rely on low-throughput and pricey experimental studies. Computational prediction models offer a promising alternative to accelerate this process. Graph neural networks have shown significant progress in protein research, but capturing long-distance structural correlations and identifying key residues in protein graphs remains challenging.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>In the present study, we propose a novel deep learning model named Hierarchical graph transformEr with contrAstive Learning (HEAL) for protein function prediction. The core feature of HEAL is its ability to capture structural semantics using a hierarchical graph Transformer, which introduces a range of super-nodes mimicking functional motifs to interact with nodes in the protein graph. These semantic-aware super-node embeddings are then aggregated with varying emphasis to produce a graph representation. To optimize the network, we utilized graph contrastive learning as a regularization technique to maximize the similarity between different views of the graph representation. Evaluation of the PDBch test set shows that HEAL-PDB, trained on fewer data, achieves comparable performance to the recent state-of-the-art methods, such as DeepFRI. Moreover, HEAL, with the added benefit of unresolved protein structures predicted by AlphaFold2, outperforms DeepFRI by a significant margin on Fmax, AUPR, and Smin metrics on PDBch test set. Additionally, when there are no experimentally resolved structures available for the proteins of interest, HEAL can still achieve better performance on AFch test set than DeepFRI and DeepGOPlus by taking advantage of AlphaFold2 predicted structures. Finally, HEAL is capable of finding functional sites through class activation mapping.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Availability and implementation<\/jats:title>\n                  <jats:p>Implementations of our HEAL can be found at https:\/\/github.com\/ZhonghuiGu\/HEAL.<\/jats:p>\n               <\/jats:sec>","DOI":"10.1093\/bioinformatics\/btad410","type":"journal-article","created":{"date-parts":[[2023,6,27]],"date-time":"2023-06-27T20:53:49Z","timestamp":1687899229000},"source":"Crossref","is-referenced-by-count":84,"title":["Hierarchical graph transformer with contrastive learning for protein function prediction"],"prefix":"10.1093","volume":"39","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-6142-4358","authenticated-orcid":false,"given":"Zhonghui","family":"Gu","sequence":"first","affiliation":[{"name":"Peking-Tsinghua Center for Life Sciences, Academy for Advanced Interdisciplinary Studies, Peking University , Beijing 100871, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiao","family":"Luo","sequence":"additional","affiliation":[{"name":"Department of Computer Science, University of California, Los Angeles , Los Angeles, CA 90024, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jiaxiao","family":"Chen","sequence":"additional","affiliation":[{"name":"Center for Quantitative Biology, Academy for Advanced Interdisciplinary Studies, Peking University , Beijing 100871, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Minghua","family":"Deng","sequence":"additional","affiliation":[{"name":"Center for Quantitative Biology, Academy for Advanced Interdisciplinary Studies, Peking University , Beijing 100871, China"},{"name":"School of Mathematics Sciences, Peking University , Beijing 100871, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8343-7587","authenticated-orcid":false,"given":"Luhua","family":"Lai","sequence":"additional","affiliation":[{"name":"Peking-Tsinghua Center for Life Sciences, Academy for Advanced Interdisciplinary Studies, Peking University , Beijing 100871, China"},{"name":"Center for Quantitative Biology, Academy for Advanced Interdisciplinary Studies, Peking University , Beijing 100871, China"},{"name":"BNLMS, College of Chemistry and Molecular Engineering, Peking University , Beijing 100871, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"286","published-online":{"date-parts":[[2023,6,27]]},"reference":[{"key":"2023071305370369200_btad410-B1","doi-asserted-by":"crossref","first-page":"403","DOI":"10.1016\/S0022-2836(05)80360-2","article-title":"Basic local alignment search tool","volume":"215","author":"Altschul","year":"1990","journal-title":"J Mol Biol"},{"key":"2023071305370369200_btad410-B2","doi-asserted-by":"crossref","first-page":"D115","DOI":"10.1093\/nar\/gkh131","article-title":"Uniprot: the universal protein knowledgebase","volume":"32","author":"Apweiler","year":"2004","journal-title":"Nucleic Acids Res"},{"key":"2023071305370369200_btad410-B3","author":"Baek"},{"key":"2023071305370369200_btad410-B4","doi-asserted-by":"crossref","first-page":"871","DOI":"10.1126\/science.abj8754","article-title":"Accurate prediction of protein structures and interactions using a three-track neural network","volume":"373","author":"Baek","year":"2021","journal-title":"Science"},{"key":"2023071305370369200_btad410-B5","doi-asserted-by":"crossref","first-page":"235","DOI":"10.1093\/nar\/28.1.235","article-title":"The protein data bank","volume":"28","author":"Berman","year":"2000","journal-title":"Nucleic Acids Res"},{"key":"2023071305370369200_btad410-B6","doi-asserted-by":"crossref","first-page":"59","DOI":"10.1038\/nmeth.3176","article-title":"Fast and sensitive protein alignment using diamond","volume":"12","author":"Buchfink","year":"2015","journal-title":"Nat Methods"},{"key":"2023071305370369200_btad410-B7","doi-asserted-by":"crossref","first-page":"2825","DOI":"10.1093\/bioinformatics\/btab198","article-title":"Tale: transformer-based protein function annotation with joint sequence\u2013label embedding","volume":"37","author":"Cao","year":"2021","journal-title":"Bioinformatics"},{"key":"2023071305370369200_btad410-B8","doi-asserted-by":"crossref","first-page":"540","DOI":"10.1016\/j.cels.2016.10.017","article-title":"Compact integration of multi-network topology for functional analysis of genes","volume":"3","author":"Cho","year":"2016","journal-title":"Cell Syst"},{"key":"2023071305370369200_btad410-B9","doi-asserted-by":"crossref","first-page":"D482","DOI":"10.1093\/nar\/gky1114","article-title":"Sifts: updated structure integration with function, taxonomy and sequences resource allows 40-fold increase in coverage of structure-based annotations for proteins","volume":"47","author":"Dana","year":"2019","journal-title":"Nucleic Acids Res"},{"key":"2023071305370369200_btad410-B10","doi-asserted-by":"crossref","first-page":"3460","DOI":"10.1093\/bioinformatics\/btv398","article-title":"Functional classification of CATH superfamilies: a domain-based approach for protein function annotation","volume":"31","author":"Das","year":"2015","journal-title":"Bioinformatics"},{"key":"2023071305370369200_btad410-B11","doi-asserted-by":"crossref","first-page":"e0198216","DOI":"10.1371\/journal.pone.0198216","article-title":"Predicting human protein function with multi-task deep neural networks","volume":"13","author":"Fa","year":"2018","journal-title":"PLoS One"},{"key":"2023071305370369200_btad410-B12","author":"Fey"},{"key":"2023071305370369200_btad410-B13","doi-asserted-by":"crossref","first-page":"e2104878118","DOI":"10.1073\/pnas.2104878118","article-title":"Neural networks to learn protein sequence\u2013function relationships from deep mutational scanning data","volume":"118","author":"Gelman","year":"2021","journal-title":"Proc Natl Acad Sci USA"},{"key":"2023071305370369200_btad410-B14","first-page":"1263","author":"Gilmer"},{"key":"2023071305370369200_btad410-B15","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1038\/s41467-021-23303-9","article-title":"Structure-based protein function prediction using graph convolutional networks","volume":"12","author":"Gligorijevi\u0107","year":"2021","journal-title":"Nat Commun"},{"key":"2023071305370369200_btad410-B16","first-page":"9729","author":"He"},{"key":"2023071305370369200_btad410-B17","doi-asserted-by":"crossref","first-page":"583","DOI":"10.1038\/s41586-021-03819-2","article-title":"Highly accurate protein structure prediction with AlphaFold","volume":"596","author":"Jumper","year":"2021","journal-title":"Nature"},{"key":"2023071305370369200_btad410-B18","author":"Kingma"},{"key":"2023071305370369200_btad410-B19","author":"Kipf"},{"key":"2023071305370369200_btad410-B20","doi-asserted-by":"crossref","first-page":"1187","DOI":"10.1093\/bioinformatics\/btaa763","article-title":"Deepgoplus: improved protein function prediction from sequence","volume":"37","author":"Kulmanov","year":"2021","journal-title":"Bioinformatics"},{"key":"2023071305370369200_btad410-B21","doi-asserted-by":"crossref","first-page":"660","DOI":"10.1093\/bioinformatics\/btx624","article-title":"Deepgo: predicting protein functions from sequence and interactions using a deep ontology-aware classifier","volume":"34","author":"Kulmanov","year":"2018","journal-title":"Bioinformatics"},{"key":"2023071305370369200_btad410-B22","doi-asserted-by":"crossref","first-page":"461","DOI":"10.1007\/s13735-022-00245-6","article-title":"Contrastive self-supervised learning: review, progress, challenges and future research directions","volume":"11","author":"Kumar","year":"2022","journal-title":"Int J Multimed Inf Retr"},{"key":"2023071305370369200_btad410-B23","doi-asserted-by":"crossref","first-page":"bbab502","DOI":"10.1093\/bib\/bbab502","article-title":"Accurate protein function prediction via graph attention networks with predicted structure information","volume":"23","author":"Lai","year":"2022","journal-title":"Brief Bioinformatics"},{"key":"2023071305370369200_btad410-B24","doi-asserted-by":"crossref","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":"2023071305370369200_btad410-B25","doi-asserted-by":"crossref","first-page":"S4","DOI":"10.1186\/gb-2008-9-s1-s4","article-title":"Genemania: a real-time multiple association network integration algorithm for predicting gene function","volume":"9","author":"Mostafavi","year":"2008","journal-title":"Genome Biol"},{"key":"2023071305370369200_btad410-B26","first-page":"32","article-title":"Pytorch: an imperative style, high-performance deep learning library","volume":"32","author":"Paszke","year":"2019","journal-title":"Adv Neural Inf Process Syst"},{"key":"2023071305370369200_btad410-B27","doi-asserted-by":"crossref","first-page":"221","DOI":"10.1038\/nmeth.2340","article-title":"A large-scale evaluation of computational protein function prediction","volume":"10","author":"Radivojac","year":"2013","journal-title":"Nat Methods"},{"key":"2023071305370369200_btad410-B28","doi-asserted-by":"crossref","first-page":"e2016239118","DOI":"10.1073\/pnas.2016239118","article-title":"Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences","volume":"118","author":"Rives","year":"2021","journal-title":"Proc Natl Acad Sci USA"},{"key":"2023071305370369200_btad410-B29","volume-title":"Proceedings of the IEEE International Conference on Computer Vision","author":"Selvaraju"},{"key":"2023071305370369200_btad410-B30","doi-asserted-by":"crossref","first-page":"2542","DOI":"10.1038\/s41467-018-04964-5","article-title":"Clustering huge protein sequence sets in linear time","volume":"9","author":"Steinegger","year":"2018","journal-title":"Nat Commun"},{"key":"2023071305370369200_btad410-B31","doi-asserted-by":"crossref","first-page":"D506","DOI":"10.1093\/nar\/gky1049","article-title":"UniProt: a worldwide hub of protein knowledge","volume":"47","author":"UniProt Consortium","year":"2019","journal-title":"Nucleic Acids Res"},{"key":"2023071305370369200_btad410-B32","doi-asserted-by":"crossref","first-page":"D439","DOI":"10.1093\/nar\/gkab1061","article-title":"AlphaFold protein structure database: massively expanding the structural coverage of protein-sequence space with high-accuracy models","volume":"50","author":"Varadi","year":"2022","journal-title":"Nucleic Acids Res"},{"key":"2023071305370369200_btad410-B33","first-page":"1","volume-title":"Advances in neural information processing systems","author":"Vaswani"},{"key":"2023071305370369200_btad410-B34","doi-asserted-by":"crossref","DOI":"10.1016\/j.gpb.2023.04.001","article-title":"NetGO 3.0: protein language model improves large-scale functional annotations","author":"Wang","year":"2023","journal-title":"Genomics Proteomics Bioinformatics"},{"key":"2023071305370369200_btad410-B35","doi-asserted-by":"crossref","first-page":"W296","DOI":"10.1093\/nar\/gky427","article-title":"Swiss-model: homology modelling of protein structures and complexes","volume":"46","author":"Waterhouse","year":"2018","journal-title":"Nucleic Acids Res"},{"key":"2023071305370369200_btad410-B36","doi-asserted-by":"crossref","first-page":"D1096","DOI":"10.1093\/nar\/gks966","article-title":"BioLiP: a semi-manually curated database for biologically relevant ligand\u2013protein interactions","volume":"41","author":"Yang","year":"2013","journal-title":"Nucleic Acids Res"},{"key":"2023071305370369200_btad410-B37","doi-asserted-by":"crossref","first-page":"7","DOI":"10.1038\/nmeth.3213","article-title":"The I-TASSER suite: protein structure and function prediction","volume":"12","author":"Yang","year":"2015","journal-title":"Nat Methods"},{"key":"2023071305370369200_btad410-B38","doi-asserted-by":"crossref","first-page":"W469","DOI":"10.1093\/nar\/gkab398","article-title":"Netgo 2.0: improving large-scale protein function prediction with massive sequence, text, domain, family and network information","volume":"49","author":"Yao","year":"2021","journal-title":"Nucleic Acids Res"},{"key":"2023071305370369200_btad410-B39","doi-asserted-by":"crossref","first-page":"i262","DOI":"10.1093\/bioinformatics\/btab270","article-title":"Deepgraphgo: graph neural network for large-scale, multispecies protein function prediction","volume":"37","author":"You","year":"2021","journal-title":"Bioinformatics"},{"key":"2023071305370369200_btad410-B40","first-page":"5812","author":"You"},{"key":"2023071305370369200_btad410-B41","author":"Yu"},{"key":"2023071305370369200_btad410-B42","first-page":"10824","volume-title":"Proceedings of the AAAI Conference on Artificial Intelligence","author":"Zeng"},{"key":"2023071305370369200_btad410-B43","doi-asserted-by":"crossref","first-page":"1900019","DOI":"10.1002\/pmic.201900019","article-title":"Deepfunc: a deep learning framework for accurate prediction of protein functions from protein sequences and interactions","volume":"19","author":"Zhang","year":"2019","journal-title":"Proteomics"},{"key":"2023071305370369200_btad410-B44","doi-asserted-by":"crossref","first-page":"lqac004","DOI":"10.1093\/nargab\/lqac004","article-title":"Panda2: protein function prediction using graph neural networks","volume":"4","author":"Zhao","year":"2022","journal-title":"NAR Genom Bioinform"},{"key":"2023071305370369200_btad410-B45","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1186\/s13059-019-1835-8","article-title":"The CAFA challenge reports improved protein function prediction and new functional annotations for hundreds of genes through experimental screens","volume":"20","author":"Zhou","year":"2019","journal-title":"Genome Biol"},{"key":"2023071305370369200_btad410-B46","doi-asserted-by":"crossref","first-page":"e1010793","DOI":"10.1371\/journal.pcbi.1010793","article-title":"Integrating unsupervised language model with triplet neural networks for protein gene ontology prediction","volume":"18","author":"Zhu","year":"2022","journal-title":"PLoS Comput Biol"}],"container-title":["Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bioinformatics\/advance-article-pdf\/doi\/10.1093\/bioinformatics\/btad410\/50725438\/btad410.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/39\/7\/btad410\/50874760\/btad410.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/39\/7\/btad410\/50874760\/btad410.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,7,13]],"date-time":"2023-07-13T06:07:39Z","timestamp":1689228459000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article\/doi\/10.1093\/bioinformatics\/btad410\/7208864"}},"subtitle":[],"editor":[{"given":"Peter","family":"Robinson","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2023,6,27]]},"references-count":46,"journal-issue":{"issue":"7","published-print":{"date-parts":[[2023,7,1]]}},"URL":"https:\/\/doi.org\/10.1093\/bioinformatics\/btad410","relation":{},"ISSN":["1367-4811"],"issn-type":[{"value":"1367-4811","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2023,7,1]]},"published":{"date-parts":[[2023,6,27]]},"article-number":"btad410"}}