{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,13]],"date-time":"2026-05-13T07:48:38Z","timestamp":1778658518098,"version":"3.51.4"},"reference-count":53,"publisher":"Oxford University Press (OUP)","issue":"17","license":[{"start":{"date-parts":[[2018,3,26]],"date-time":"2018-03-26T00:00:00Z","timestamp":1522022400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/academic.oup.com\/journals\/pages\/open_access\/funder_policies\/chorus\/standard_publication_model"}],"funder":[{"name":"Intramural Research Programs"},{"DOI":"10.13039\/100000002","name":"National Institutes of Health","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100000002","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000092","name":"National Library of Medicine","doi-asserted-by":"crossref","id":[{"id":"10.13039\/100000092","id-type":"DOI","asserted-by":"crossref"}]},{"name":"NCBI"},{"DOI":"10.13039\/100000002","name":"NIH","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100000002","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Ramakanth Kavuluru and Anthony Rios"},{"DOI":"10.13039\/100000092","name":"National Library of Medicine","doi-asserted-by":"crossref","award":["R21LM012274"],"award-info":[{"award-number":["R21LM012274"]}],"id":[{"id":"10.13039\/100000092","id-type":"DOI","asserted-by":"crossref"}]},{"name":"NVIDIA Corporation"},{"name":"Titan X Pascal GPU"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2018,9,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:sec>\n                  <jats:title>Motivation<\/jats:title>\n                  <jats:p>Creating large datasets for biomedical relation classification can be prohibitively expensive. While some datasets have been curated to extract protein\u2013protein and drug\u2013drug interactions (PPIs and DDIs) from text, we are also interested in other interactions including gene\u2013disease and chemical\u2013protein connections. Also, many biomedical researchers have begun to explore ternary relationships. Even when annotated data are available, many datasets used for relation classification are inherently biased. For example, issues such as sample selection bias typically prevent models from generalizing in the wild. To address the problem of cross-corpora generalization, we present a novel adversarial learning algorithm for unsupervised domain adaptation tasks where no labeled data are available in the target domain. Instead, our method takes advantage of unlabeled data to improve biased classifiers through learning domain-invariant features via an adversarial process. Finally, our method is built upon recent advances in neural network (NN) methods.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>We experiment by extracting PPIs and DDIs from text. In our experiments, we show domain invariant features can be learned in NNs such that classifiers trained for one interaction type (protein\u2013protein) can be re-purposed to others (drug\u2013drug). We also show that our method can adapt to different source and target pairs of PPI datasets. Compared to prior convolutional and recurrent NN-based relation classification methods without domain adaptation, we achieve improvements as high as 30% in F1-score. Likewise, we show improvements over state-of-the-art adversarial methods.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Availability and implementation<\/jats:title>\n                  <jats:p>Experimental code is available at https:\/\/github.com\/bionlproc\/adversarial-relation-classification.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Supplementary information<\/jats:title>\n                  <jats:p>Supplementary data are available at Bioinformatics online.<\/jats:p>\n               <\/jats:sec>","DOI":"10.1093\/bioinformatics\/bty190","type":"journal-article","created":{"date-parts":[[2018,3,25]],"date-time":"2018-03-25T19:06:51Z","timestamp":1522004811000},"page":"2973-2981","source":"Crossref","is-referenced-by-count":27,"title":["Generalizing biomedical relation classification with neural adversarial domain adaptation"],"prefix":"10.1093","volume":"34","author":[{"given":"Anthony","family":"Rios","sequence":"first","affiliation":[{"name":"National Library of Medicine (NLM), National Center for Biotechnology Information (NCBI), National Institutes of Health (NIH), Bethesda, MD, USA"},{"name":"Department of Computer Science, University of Kentucky, Lexington, KY, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ramakanth","family":"Kavuluru","sequence":"additional","affiliation":[{"name":"Department of Computer Science, University of Kentucky, Lexington, KY, USA"},{"name":"Division of Biomedical Informatics, Department of Internal Medicine, Lexington, KY, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zhiyong","family":"Lu","sequence":"additional","affiliation":[{"name":"National Library of Medicine (NLM), National Center for Biotechnology Information (NCBI), National Institutes of Health (NIH), Bethesda, MD, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"286","published-online":{"date-parts":[[2018,3,26]]},"reference":[{"key":"2023061313375918400_bty190-B1","author":"Asada","year":"2017"},{"key":"2023061313375918400_bty190-B2","doi-asserted-by":"crossref","first-page":"206","DOI":"10.1016\/j.jbi.2015.04.006","article-title":"Toward a complete dataset of drug\u2013drug interaction information from publicly available sources","volume":"55","author":"Ayvaz","year":"2015","journal-title":"J. Biomed. Informat"},{"key":"2023061313375918400_bty190-B3","first-page":"1052","author":"Bethard","year":"2017"},{"key":"2023061313375918400_bty190-B4","doi-asserted-by":"crossref","first-page":"139","DOI":"10.1016\/j.artmed.2004.07.016","article-title":"Comparative experiments on learning information extractors for proteins and their interactions","volume":"33","author":"Bunescu","year":"2005","journal-title":"Artif. Intell. Med"},{"key":"2023061313375918400_bty190-B5","author":"Bunescu","year":"2005"},{"key":"2023061313375918400_bty190-B6","first-page":"767","author":"Chen","year":"2012"},{"key":"2023061313375918400_bty190-B7","author":"Council","year":"2009"},{"key":"2023061313375918400_bty190-B8","first-page":"256","author":"Daum\u00e9","year":"2007"},{"key":"2023061313375918400_bty190-B9","author":"dos Santos","year":"2015"},{"key":"2023061313375918400_bty190-B10","author":"Ganin","year":"2015"},{"key":"2023061313375918400_bty190-B11","first-page":"1","article-title":"Domain-adversarial training of neural networks","volume":"17","author":"Ganin","year":"2016","journal-title":"JMLR"},{"key":"2023061313375918400_bty190-B12","doi-asserted-by":"crossref","first-page":"2451","DOI":"10.1162\/089976600300015015","article-title":"Learning to forget: continual prediction with lstm","volume":"12","author":"Gers","year":"2000","journal-title":"Neural Comput"},{"key":"2023061313375918400_bty190-B13","author":"Glorot","year":"2011"},{"key":"2023061313375918400_bty190-B14","author":"Glorot","year":"2011"},{"key":"2023061313375918400_bty190-B15","author":"Gong","year":"2012"},{"key":"2023061313375918400_bty190-B16","author":"Goodfellow","year":"2016"},{"key":"2023061313375918400_bty190-B17","author":"Goodfellow","year":"2014"},{"key":"2023061313375918400_bty190-B18","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-642-24797-2","volume-title":"Supervised Sequence Labelling with Recurrent Neural Networks. Vol. 385","author":"Graves","year":"2012"},{"key":"2023061313375918400_bty190-B19","author":"Guo","year":"2017"},{"key":"2023061313375918400_bty190-B20","author":"Huang","year":"2006"},{"key":"2023061313375918400_bty190-B21","author":"Kavuluru","year":"2017"},{"key":"2023061313375918400_bty190-B22","author":"Kingma","year":"2015"},{"key":"2023061313375918400_bty190-B23","doi-asserted-by":"crossref","first-page":"733.","DOI":"10.1038\/nrg2825","article-title":"Tackling the widespread and critical impact of batch effects in high-throughput data","volume":"11","author":"Leek","year":"2010","journal-title":"Nat. Rev. Genet"},{"key":"2023061313375918400_bty190-B24","doi-asserted-by":"crossref","first-page":"1.","DOI":"10.1155\/2016\/4310634","article-title":"Drug\u2013drug interaction extraction via convolutional neural networks","volume":"2016","author":"Liu","year":"2016","journal-title":"Comput. Math. Methods Med"},{"key":"2023061313375918400_bty190-B25","author":"Long","year":"2015"},{"key":"2023061313375918400_bty190-B26","first-page":"136","author":"Long","year":"2016"},{"key":"2023061313375918400_bty190-B27","author":"Matos","year":"2017"},{"key":"2023061313375918400_bty190-B28","author":"Mikolov","year":"2013"},{"key":"2023061313375918400_bty190-B29","author":"Ming Harry Hsu","year":"2015"},{"key":"2023061313375918400_bty190-B30","author":"Mou","year":"2016"},{"key":"2023061313375918400_bty190-B31","author":"Nair","year":"2010"},{"key":"2023061313375918400_bty190-B32","first-page":"68","author":"Nguyen","year":"2014"},{"key":"2023061313375918400_bty190-B33","author":"Nguyen","year":"2015"},{"key":"2023061313375918400_bty190-B34","author":"Nguyen","year":"2015"},{"key":"2023061313375918400_bty190-B35","doi-asserted-by":"crossref","first-page":"8.","DOI":"10.1186\/1751-0473-5-8","article-title":"Open source tool for prediction of genome wide protein-protein interaction network based on ortholog information","volume":"5","author":"Pedamallu","year":"2010","journal-title":"Source Code Biol. Med"},{"key":"2023061313375918400_bty190-B36","author":"Peng","year":"2017"},{"key":"2023061313375918400_bty190-B37","author":"Peng","year":"2017"},{"key":"2023061313375918400_bty190-B38","author":"Plank","year":"2013"},{"key":"2023061313375918400_bty190-B39","doi-asserted-by":"crossref","first-page":"50.","DOI":"10.1186\/1471-2105-8-50","article-title":"Bioinfer: a corpus for information extraction in the biomedical domain","volume":"8","author":"Pyysalo","year":"2007","journal-title":"BMC Bioinformatics"},{"key":"2023061313375918400_bty190-B40","author":"Rink","year":"2010"},{"key":"2023061313375918400_bty190-B41","author":"Salimans","year":"2016"},{"key":"2023061313375918400_bty190-B42","doi-asserted-by":"crossref","first-page":"152","DOI":"10.1016\/j.jbi.2014.05.007","article-title":"Lessons learnt from the ddiextraction-2013 shared task","volume":"51","author":"Segura-Bedmar","year":"2014","journal-title":"J. Biomed. Informat"},{"key":"2023061313375918400_bty190-B43","doi-asserted-by":"crossref","first-page":"2539","DOI":"10.1093\/bioinformatics\/btx196","article-title":"Removal of batch effects using distribution-matching residual networks","volume":"33","author":"Shaham","year":"2017","journal-title":"Bioinformatics"},{"key":"2023061313375918400_bty190-B44","doi-asserted-by":"crossref","first-page":"e1005017.","DOI":"10.1371\/journal.pcbi.1005017","article-title":"Text mining genotype-phenotype relationships from biomedical literature for database curation and precision medicine","volume":"12","author":"Singhal","year":"2016","journal-title":"PLoS Comput. Biol"},{"key":"2023061313375918400_bty190-B45","author":"Szegedy","year":"2016"},{"key":"2023061313375918400_bty190-B46","first-page":"2962","author":"Tzeng","year":"2017"},{"key":"2023061313375918400_bty190-B47","author":"Vu","year":"2016"},{"key":"2023061313375918400_bty190-B48","first-page":"1.","article-title":"Gnormplus: an integrative approach for tagging genes, gene families, and protein domains","volume":"2015","author":"Wei","year":"2015","journal-title":"BioMed. Res. Int"},{"key":"2023061313375918400_bty190-B49","author":"Xu","year":"2015"},{"key":"2023061313375918400_bty190-B50","author":"Zeng","year":"2014"},{"key":"2023061313375918400_bty190-B51","doi-asserted-by":"crossref","first-page":"828","DOI":"10.1093\/bioinformatics\/btx659","article-title":"Drug\u2013drug interaction extraction via hierarchical RNNS on sequence and shortest dependency paths","volume":"34","author":"Zhang","year":"2017","journal-title":"Bioinformatics"},{"key":"2023061313375918400_bty190-B52","doi-asserted-by":"crossref","first-page":"3444","DOI":"10.1093\/bioinformatics\/btw486","article-title":"Drug drug interaction extraction from biomedical literature using syntax convolutional neural network","volume":"32","author":"Zhao","year":"2016","journal-title":"Bioinformatics"},{"key":"2023061313375918400_bty190-B53","author":"Zhou","year":"2016"}],"container-title":["Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/34\/17\/2973\/50581920\/bioinformatics_34_17_2973.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/34\/17\/2973\/50581920\/bioinformatics_34_17_2973.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,6,13]],"date-time":"2023-06-13T13:39:37Z","timestamp":1686663577000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article\/34\/17\/2973\/4953706"}},"subtitle":[],"editor":[{"given":"Jonathan","family":"Wren","sequence":"additional","affiliation":[],"role":[{"role":"editor","vocabulary":"crossref"}]}],"short-title":[],"issued":{"date-parts":[[2018,3,26]]},"references-count":53,"journal-issue":{"issue":"17","published-print":{"date-parts":[[2018,9,1]]}},"URL":"https:\/\/doi.org\/10.1093\/bioinformatics\/bty190","relation":{},"ISSN":["1367-4803","1367-4811"],"issn-type":[{"value":"1367-4803","type":"print"},{"value":"1367-4811","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2018,9,1]]},"published":{"date-parts":[[2018,3,26]]}}}