{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T06:16:16Z","timestamp":1792736176207,"version":"4.5.0"},"reference-count":61,"publisher":"Oxford University Press (OUP)","issue":"2","license":[{"start":{"date-parts":[[2025,9,30]],"date-time":"2025-09-30T00:00:00Z","timestamp":1759190400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"funder":[{"DOI":"10.13039\/100000038","name":"US Food and Drug Administration","doi-asserted-by":"crossref","id":[{"id":"10.13039\/100000038","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,12,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Computational toxicology plays an important role in risk assessment and drug safety. The field has been traditionally dominated by Quantitative Structure\u2013Activity Relationships (QSARs), which predict toxicological effects based solely on chemical structure. Although QSARs have achieved successes, their structure reliance limits drug toxicity predictions, where small structural modifications may cause major toxicity changes. Advances in artificial intelligence (AI), especially text embedding and generative AI, provide an opportunity to enhance toxicity predictions by leveraging broader chemical knowledge and its integration with structural data. In this study, we propose a novel framework, Quantitative Knowledge\u2013Activity Relationships (QKARs), which predicts toxicity using domain-specific knowledge. We developed QKAR models for two drug toxicity endpoints, drug-induced liver injury (DILI) and drug-induced cardiotoxicity (DICT), using three different knowledge representations with varying levels of knowledge. The representations based on comprehensive knowledge of the drugs yielded better prediction than those with simpler knowledge. Five machine learning algorithms of distinct complexity were applied in QKAR models, and we observed little association between model complexity and performance. Further, we evaluated QKARs against QSARs on the same endpoints using identical datasets. We found that QKARs consistently outperformed QSARs for DILI and DICT. Notably, QKARs demonstrated better capability than QSARs in differentiating drugs with similar structures but different liver toxicity profiles. We also investigated integrating knowledge-based and structure-based representations, Q(K\u2009+\u2009S)ARs, for further enhanced prediction accuracy. Our findings demonstrate the potential of QKARs as a robust alternative to QSARs, offering additional opportunities in drug toxicity assessments by leveraging both domain-specific knowledge and structural data.<\/jats:p>","DOI":"10.1093\/toxsci\/kfaf135","type":"journal-article","created":{"date-parts":[[2025,9,28]],"date-time":"2025-09-28T11:48:07Z","timestamp":1759060087000},"page":"269-278","source":"Crossref","is-referenced-by-count":7,"title":["Beyond QSARs: Quantitative Knowledge\u2013Activity Relationships (QKARs) for enhanced drug toxicity prediction"],"prefix":"10.1093","volume":"208","author":[{"given":"Ting","family":"Li","sequence":"first","affiliation":[{"name":"National Center for Toxicological Research, US Food and Drug Administration , Jefferson, AR 72079,","place":["United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yanyan","family":"Qu","sequence":"additional","affiliation":[{"name":"National Center for Toxicological Research, US Food and Drug Administration , Jefferson, AR 72079,","place":["United States"]},{"name":"University of Arkansas at Little Rock and University of Arkansas for Medical Sciences Joint Bioinformatics Program , Little Rock, AR 72204,","place":["United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alexander","family":"Chen","sequence":"additional","affiliation":[{"name":"National Center for Toxicological Research, US Food and Drug Administration , Jefferson, AR 72079,","place":["United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shraddha","family":"Thakkar","sequence":"additional","affiliation":[{"name":"Office of Translational Sciences, Center for Drug Evaluation and Research, US Food and Drug Administration , Silver Spring, MD 20903,","place":["United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0721-2500","authenticated-orcid":false,"given":"Dongying","family":"Li","sequence":"additional","affiliation":[{"name":"National Center for Toxicological Research, US Food and Drug Administration , Jefferson, AR 72079,","place":["United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3488-6148","authenticated-orcid":false,"given":"Weida","family":"Tong","sequence":"additional","affiliation":[{"name":"National Center for Toxicological Research, US Food and Drug Administration , Jefferson, AR 72079,","place":["United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"286","published-online":{"date-parts":[[2025,9,30]]},"reference":[{"key":"2025112517214735200_kfaf135-B1","author":"Bang","year":"2023"},{"key":"2025112517214735200_kfaf135-B2","doi-asserted-by":"crossref","first-page":"868","DOI":"10.1007\/s10439-023-03172-7","article-title":"Role of chat GPT in public health","volume":"51","author":"Biswas","year":"2023","journal-title":"Ann Biomed Eng"},{"key":"2025112517214735200_kfaf135-B3","author":"Cao","year":"2023"},{"key":"2025112517214735200_kfaf135-B4","doi-asserted-by":"crossref","first-page":"17","DOI":"10.3389\/frai.2019.00017","article-title":"Descriptor free qsar modeling using deep learning with long short-term memory neural networks","volume":"2","author":"Chakravarti","year":"2019","journal-title":"Front Artif Intell"},{"key":"2025112517214735200_kfaf135-B5","doi-asserted-by":"crossref","first-page":"93","DOI":"10.1186\/s13321-021-00570-8","article-title":"Chemical toxicity prediction based on semi-supervised learning and graph convolutional neural network","volume":"13","author":"Chen","year":"2021","journal-title":"J Cheminform"},{"key":"2025112517214735200_kfaf135-B6","doi-asserted-by":"crossref","first-page":"7141","DOI":"10.1038\/s41467-023-42933-9","article-title":"A generative adversarial network model alternative to animal studies for clinical pathology assessment","volume":"14","author":"Chen","year":"2023","journal-title":"Nat Commun"},{"key":"2025112517214735200_kfaf135-B7","doi-asserted-by":"crossref","first-page":"4977","DOI":"10.1021\/jm4004285","article-title":"Qsar modeling: where have you been? Where are you going to?","volume":"57","author":"Cherkasov","year":"2014","journal-title":"J Med Chem"},{"key":"2025112517214735200_kfaf135-B8","author":"Chithrananda","year":"2020"},{"key":"2025112517214735200_kfaf135-B9","doi-asserted-by":"crossref","first-page":"1273","DOI":"10.1021\/ci010132r","article-title":"Reoptimization of mdl keys for use in drug discovery","volume":"42","author":"Durant","year":"2002","journal-title":"J Chem Inf Comput Sci"},{"key":"2025112517214735200_kfaf135-B10","first-page":"2224","article-title":"Convolutional networks on graphs for learning molecular fingerprints","author":"Duvenaud","year":"2015"},{"key":"2025112517214735200_kfaf135-B11","author":"Fabian","year":"2020"},{"key":"2025112517214735200_kfaf135-B12","doi-asserted-by":"crossref","first-page":"1189","DOI":"10.1021\/ci100176x","article-title":"Trust, but verify: on the importance of chemical structure curation in cheminformatics and qsar modeling research","volume":"50","author":"Fourches","year":"2010","journal-title":"J Chem Inf Model"},{"key":"2025112517214735200_kfaf135-B13","first-page":"277","article-title":"Generative AI and ChatGPT: applications, challenges, and AI-human collaboration","volume":"25","author":"Fui-Hoon Nah","year":"2023","journal-title":"J Inf Technol Case Appl Res"},{"key":"2025112517214735200_kfaf135-B14","author":"Goh","year":"2017"},{"key":"2025112517214735200_kfaf135-B15","doi-asserted-by":"crossref","first-page":"213","DOI":"10.1002\/pds.1207","article-title":"A systematic review of NSAIDs withdrawn from the market due to hepatotoxicity: lessons learned from the bromfenac experience","volume":"15","author":"Goldkind","year":"2006","journal-title":"Pharmacoepidemiol Drug Saf"},{"key":"2025112517214735200_kfaf135-B16","first-page":"59662","article-title":"What can large language models do in chemistry? A comprehensive benchmark on eight tasks","volume":"36","author":"Guo","year":"2023","journal-title":"Adv Neural Inf Process Syst"},{"key":"2025112517214735200_kfaf135-B17","doi-asserted-by":"crossref","first-page":"178","DOI":"10.1038\/194178b0","article-title":"Correlation of biological activity of phenoxyacetic acids with hammett substituent constants and partition coefficients","volume":"194","author":"Hansch","year":"1962","journal-title":"Nature"},{"key":"2025112517214735200_kfaf135-B18","doi-asserted-by":"crossref","first-page":"153975","DOI":"10.1016\/j.tox.2024.153975","article-title":"Predictive, integrative, and regulatory aspects of AI-driven computational toxicology\u2014highlights of the German pharm-tox summit (GPTS) 2024","volume":"509","author":"Ha\u00dfmann","year":"2024","journal-title":"Toxicology"},{"key":"2025112517214735200_kfaf135-B19","doi-asserted-by":"crossref","first-page":"526","DOI":"10.1186\/s12859-018-2523-5","article-title":"Convolutional neural network based on smiles representation of compounds for detecting chemical motif","volume":"19","author":"Hirohara","year":"2018","journal-title":"BMC Bioinformatics"},{"key":"2025112517214735200_kfaf135-B20","doi-asserted-by":"crossref","first-page":"1337","DOI":"10.1021\/ci800038f","article-title":"Mold2, molecular descriptors from 2D structures for chemoinformatics and toxicoinformatics","volume":"48","author":"Hong","year":"2008","journal-title":"J Chem Inf Model"},{"key":"2025112517214735200_kfaf135-B21","doi-asserted-by":"crossref","first-page":"1283","DOI":"10.1007\/s11030-021-10250-2","article-title":"QSAR modeling without descriptors using graph convolutional neural networks: the case of mutagenicity prediction","volume":"25","author":"Hung","year":"2021","journal-title":"Mol Divers"},{"key":"2025112517214735200_kfaf135-B22","doi-asserted-by":"crossref","first-page":"015022","DOI":"10.1088\/2632-2153\/ac3ffb","article-title":"Chemformer: a pre-trained transformer for computational chemistry","volume":"3","author":"Irwin","year":"2022","journal-title":"Mach Learn Sci Technol"},{"key":"2025112517214735200_kfaf135-B23","doi-asserted-by":"crossref","first-page":"14","DOI":"10.1093\/toxsci\/kfm297","article-title":"Computational toxicology\u2014a state of the science mini review","volume":"103","author":"Kavlock","year":"2008","journal-title":"Toxicol Sci"},{"key":"2025112517214735200_kfaf135-B24","author":"Kusupati","year":"2022"},{"key":"2025112517214735200_kfaf135-B25","doi-asserted-by":"crossref","first-page":"310","DOI":"10.1038\/s41746-024-01317-z","article-title":"Bridging organ transcriptomics for advancing multiple organ toxicity assessment with a generative ai approach","volume":"7","author":"Li","year":"2024","journal-title":"NPJ Digit Med"},{"key":"2025112517214735200_kfaf135-B26","doi-asserted-by":"crossref","first-page":"105486","DOI":"10.1016\/j.yrtph.2023.105486","article-title":"DeepAmes: a deep learning-powered Ames test predictive model with potential for regulatory application","volume":"144","author":"Li","year":"2023","journal-title":"Regul Toxicol Pharmacol"},{"key":"2025112517214735200_kfaf135-B27","doi-asserted-by":"crossref","first-page":"550","DOI":"10.1021\/acs.chemrestox.0c00374","article-title":"Deepdili: deep learning-powered drug-induced liver injury prediction using model-level representation","volume":"34","author":"Li","year":"2021","journal-title":"Chem Res Toxicol"},{"key":"2025112517214735200_kfaf135-B28","doi-asserted-by":"crossref","first-page":"757780","DOI":"10.3389\/frai.2021.757780","article-title":"Deepcarc: deep learning-powered carcinogenicity prediction using model-level representation","volume":"4","author":"Li","year":"2021","journal-title":"Front Artif Intell"},{"key":"2025112517214735200_kfaf135-B29","doi-asserted-by":"crossref","first-page":"75","DOI":"10.1186\/s12911-024-02481-8","article-title":"Exploring the potential of chatgpt in medical dialogue summarization: a study on consistency with human preferences","volume":"24","author":"Liu","year":"2024","journal-title":"BMC Med Inform Decis Mak"},{"key":"2025112517214735200_kfaf135-B30","volume-title":"On outliers and activity cliffs why QSAR often disappoints","author":"Maggiora","year":"2006"},{"key":"2025112517214735200_kfaf135-B31","first-page":"237","article-title":"Dragon software: an easy approach to molecular descriptor calculations","volume":"56","author":"Mauri","year":"2006","journal-title":"Match"},{"key":"2025112517214735200_kfaf135-B32","first-page":"e0000205","volume-title":"ChatGPT passing usmle shines a spotlight on the flaws of medical education","author":"Mbakwe","year":"2023"},{"key":"2025112517214735200_kfaf135-B33","doi-asserted-by":"crossref","first-page":"114","DOI":"10.1038\/s41467-022-35692-6","article-title":"Leveraging molecular structure and bioactivity with chemical language models for de novo drug design","volume":"14","author":"Moret","year":"2023","journal-title":"Nat Commun"},{"key":"2025112517214735200_kfaf135-B34","doi-asserted-by":"crossref","first-page":"4","DOI":"10.1186\/s13321-018-0258-y","article-title":"Mordred: a molecular descriptor calculator","volume":"10","author":"Moriwaki","year":"2018","journal-title":"J Cheminform"},{"key":"2025112517214735200_kfaf135-B35","doi-asserted-by":"crossref","first-page":"3525","DOI":"10.1039\/D0CS00098A","article-title":"QSAR without borders","volume":"49","author":"Muratov","year":"2020","journal-title":"Chem Soc Rev"},{"key":"2025112517214735200_kfaf135-B36","doi-asserted-by":"crossref","first-page":"303","DOI":"10.1016\/j.drudis.2007.12.007","article-title":"Computational toxicology in drug development","volume":"13","author":"Muster","year":"2008","journal-title":"Drug Discov Today"},{"key":"2025112517214735200_kfaf135-B37","doi-asserted-by":"crossref","first-page":"633","DOI":"10.1517\/17460441.2010.492827","article-title":"Advances in computational methods to predict the biological activity of compounds","volume":"5","author":"Nantasenamat","year":"2010","journal-title":"Expert Opin Drug Discov"},{"key":"2025112517214735200_kfaf135-B38","doi-asserted-by":"crossref","first-page":"10","DOI":"10.1186\/s12916-016-0553-2","article-title":"Post-marketing withdrawal of 462 medicinal products because of adverse drug reactions: a systematic review of the world literature","volume":"14","author":"Onakpoya","year":"2016","journal-title":"BMC Med"},{"key":"2025112517214735200_kfaf135-B39","doi-asserted-by":"publisher","author":"P\u00e9rez Sant\u00edn","year":"2021","DOI":"10.1002\/wcms.1516"},{"key":"2025112517214735200_kfaf135-B40","doi-asserted-by":"crossref","first-page":"eaap7885","DOI":"10.1126\/sciadv.aap7885","article-title":"Deep reinforcement learning for de novo drug design","volume":"4","author":"Popova","year":"2018","journal-title":"Sci Adv"},{"key":"2025112517214735200_kfaf135-B41","doi-asserted-by":"crossref","first-page":"e202304359","DOI":"10.1002\/slct.202304359","article-title":"The future of chatgpt in medicinal chemistry: harnessing AI for accelerated drug discovery","volume":"9","author":"Pradhan","year":"2024","journal-title":"ChemistrySelect"},{"key":"2025112517214735200_kfaf135-B42","doi-asserted-by":"crossref","first-page":"103770","DOI":"10.1016\/j.drudis.2023.103770","article-title":"Dictrank: the largest reference list of 1318 human drugs ranked by risk of drug-induced cardiotoxicity using FDA labeling","volume":"28","author":"Qu","year":"2023","journal-title":"Drug Discov Today"},{"key":"2025112517214735200_kfaf135-B43","doi-asserted-by":"crossref","first-page":"647","DOI":"10.1021\/acs.chemrestox.4c00428","article-title":"Dictrank is a reliable dataset for cardiotoxicity prediction using machine learning methods","volume":"38","author":"Qu","year":"2025","journal-title":"Chem Res Toxicol"},{"key":"2025112517214735200_kfaf135-B44","doi-asserted-by":"crossref","first-page":"275","DOI":"10.1007\/s10787-009-0016-x","article-title":"Ibuprofen: pharmacology, efficacy and safety","volume":"17","author":"Rainsford","year":"2009","journal-title":"Inflammopharmacology"},{"key":"2025112517214735200_kfaf135-B45","doi-asserted-by":"crossref","first-page":"742","DOI":"10.1021\/ci100050t","article-title":"Extended-connectivity fingerprints","volume":"50","author":"Rogers","year":"2010","journal-title":"J Chem Inf Model"},{"key":"2025112517214735200_kfaf135-B46","doi-asserted-by":"crossref","first-page":"1256","DOI":"10.1038\/s42256-022-00580-7","article-title":"Large-scale chemical language representations capture molecular structure and properties","volume":"4","author":"Ross","year":"2022","journal-title":"Nat Mach Intell"},{"key":"2025112517214735200_kfaf135-B47","doi-asserted-by":"crossref","first-page":"13863","DOI":"10.1038\/s41598-022-18086-y","article-title":"Predicting odor from molecular structure: a multi-label classification approach","volume":"12","author":"Saini","year":"2022","journal-title":"Sci Rep"},{"key":"2025112517214735200_kfaf135-B48","doi-asserted-by":"crossref","first-page":"104452","DOI":"10.1016\/j.drudis.2025.104452","article-title":"New approach methodologies (NAMs) for drug-induced liver injury (DILI): where are we now?","volume":"30","author":"Shrimali","year":"2025","journal-title":"Drug Discov Today"},{"key":"2025112517214735200_kfaf135-B49","doi-asserted-by":"crossref","first-page":"1174","DOI":"10.1021\/acs.chemrestox.2c00375","article-title":"A review on the recent applications of deep learning in predictive drug toxicological studies","volume":"36","author":"Sinha","year":"2023","journal-title":"Chem Res Toxicol"},{"key":"2025112517214735200_kfaf135-B50","doi-asserted-by":"crossref","first-page":"201","DOI":"10.1016\/j.drudis.2019.09.022","article-title":"Drug-induced liver injury severity and toxicity (dilist): binary classification of 1279 drugs by human hepatotoxicity","volume":"25","author":"Thakkar","year":"2020","journal-title":"Drug Discov Today"},{"key":"2025112517214735200_kfaf135-B51","doi-asserted-by":"crossref","first-page":"408","DOI":"10.1007\/s00894-022-05373-8","article-title":"Admetboost: a web server for accurate ADMET prediction","volume":"28","author":"Tian","year":"2022","journal-title":"J Mol Model"},{"key":"2025112517214735200_kfaf135-B52","doi-asserted-by":"crossref","first-page":"141","DOI":"10.1038\/s41573-023-00832-0","article-title":"Integrating QSAR modelling and deep learning in drug discovery: the emergence of deep QSAR","volume":"23","author":"Tropsha","year":"2024","journal-title":"Nat Rev Drug Discov"},{"key":"2025112517214735200_kfaf135-B53","doi-asserted-by":"crossref","first-page":"105173","DOI":"10.1016\/j.ijmedinf.2023.105173","article-title":"Performance and exploration of ChatGPT in medical examination, records and education in Chinese: pave the way for medical AI","volume":"177","author":"Wang","year":"2023","journal-title":"Int J Med Inform"},{"key":"2025112517214735200_kfaf135-B54","doi-asserted-by":"crossref","first-page":"1315","DOI":"10.1016\/j.jmgm.2008.01.002","article-title":"The importance of the domain of applicability in QSAR modeling","volume":"26","author":"Weaver","year":"2008","journal-title":"J Mol Graph Model"},{"key":"2025112517214735200_kfaf135-B55","doi-asserted-by":"crossref","first-page":"1692","DOI":"10.1039\/C8SC04175J","article-title":"Learning continuous and data-driven molecular descriptors by translating equivalent chemical representations","volume":"10","author":"Winter","year":"2019","journal-title":"Chem Sci"},{"key":"2025112517214735200_kfaf135-B56","doi-asserted-by":"crossref","first-page":"541","DOI":"10.1021\/acs.chemrestox.0c00373","article-title":"Trade-off predictivity and explainability for machine-learning powered predictive toxicology: an in-depth investigation with tox21 data sets","volume":"34","author":"Wu","year":"2021","journal-title":"Chem Res Toxicol"},{"key":"2025112517214735200_kfaf135-B57","doi-asserted-by":"publisher","author":"Xia","year":"2022","DOI":"10.24963\/ijcai.2023\/760"},{"key":"2025112517214735200_kfaf135-B58","doi-asserted-by":"crossref","first-page":"104018","DOI":"10.1016\/j.drudis.2024.104018","article-title":"Text summarization with ChatGPT for drug labeling documents","volume":"29","author":"Ying","year":"2024","journal-title":"Drug Discov Today"},{"key":"2025112517214735200_kfaf135-B59","doi-asserted-by":"crossref","first-page":"025035","DOI":"10.1088\/2632-2153\/acdb30","article-title":"Selformer: molecular representation learning via selfies language models","volume":"4","author":"Y\u00fcksel","year":"2023","journal-title":"Mach Learn Sci Technol"},{"key":"2025112517214735200_kfaf135-B60","doi-asserted-by":"crossref","first-page":"18048","DOI":"10.1021\/jacs.3c05819","article-title":"Chatgpt chemistry assistant for text mining and the prediction of MOF synthesis","volume":"145","author":"Zheng","year":"2023","journal-title":"J Am Chem Soc"},{"key":"2025112517214735200_kfaf135-B61","doi-asserted-by":"publisher","author":"Zhou","year":"2023","DOI":"10.26434\/chemrxiv-2022-jjm0j"}],"container-title":["Toxicological Sciences"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/toxsci\/advance-article-pdf\/doi\/10.1093\/toxsci\/kfaf135\/64435671\/kfaf135.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/toxsci\/article-pdf\/208\/2\/269\/64435671\/kfaf135.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/toxsci\/article-pdf\/208\/2\/269\/64435671\/kfaf135.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,25]],"date-time":"2025-11-25T22:22:01Z","timestamp":1764109321000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/toxsci\/article\/208\/2\/269\/8268591"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,9,30]]},"references-count":61,"journal-issue":{"issue":"2","published-online":{"date-parts":[[2025,9,30]]},"published-print":{"date-parts":[[2025,12,1]]}},"URL":"https:\/\/doi.org\/10.1093\/toxsci\/kfaf135","relation":{},"ISSN":["1096-6080","1096-0929"],"issn-type":[{"value":"1096-6080","type":"print"},{"value":"1096-0929","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2025,12]]},"published":{"date-parts":[[2025,9,30]]}}}