{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,18]],"date-time":"2026-08-18T10:24:30Z","timestamp":1787048670154,"version":"build-2736575974"},"reference-count":84,"publisher":"American Chemical Society (ACS)","issue":"12","license":[{"start":{"date-parts":[[2020,12,1]],"date-time":"2020-12-01T00:00:00Z","timestamp":1606780800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2020,12,1]],"date-time":"2020-12-01T00:00:00Z","timestamp":1606780800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"},{"start":{"date-parts":[[2020,12,1]],"date-time":"2020-12-01T00:00:00Z","timestamp":1606780800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-045"}],"funder":[{"DOI":"10.13039\/100006108","name":"National Center for Advancing Translational Sciences","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100006108","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2020,12,28]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>The rise of novel artificial intelligence (AI) methods necessitates their benchmarking against classical machine learning for a typical drug-discovery project. Inhibition of the potassium ion channel, whose alpha subunit is encoded by the human ether-a\u0300-go-go-related gene (hERG), leads to a prolonged QT interval of the cardiac action potential and is a significant safety pharmacology target for the development of new medicines. Several computational approaches have been employed to develop prediction models for the assessment of hERG liabilities of small molecules including recent work using deep learning methods. Here, we perform a comprehensive comparison of hERG effect prediction models based on classical approaches (random forests and gradient boosting) and modern AI methods [deep neural networks (DNNs) and recurrent neural networks (RNNs)]. The training set (\u223c9000 compounds) was compiled by integrating the hERG bioactivity data from the ChEMBL database with experimental data generated from an in-house, high-throughput thallium flux assay. We utilized different molecular descriptors including the latent descriptors, which are real-value continuous vectors derived from chemical autoencoders trained on a large chemical space (&amp;gt;1.5 million compounds). The models were prospectively validated on \u223c840 in-house compounds screened in the same thallium flux assay. The best results were obtained with the XGBoost method and RDKit descriptors. The comparison of models based only on latent descriptors revealed that the DNNs performed significantly better than the classical methods. The RNNs that operate on SMILES provided the highest model sensitivity. The best models were merged into a consensus model that offered superior performance compared to reference models from academic and commercial domains. Furthermore, we shed light on the potential of AI methods to exploit the big data in chemistry and generate novel chemical representations useful in predictive modeling and tailoring a new chemical space.<\/jats:p>","DOI":"10.1021\/acs.jcim.0c00884","type":"journal-article","created":{"date-parts":[[2020,12,1]],"date-time":"2020-12-01T17:30:12Z","timestamp":1606843812000},"page":"6007-6019","source":"Crossref","is-referenced-by-count":64,"title":["Critical Assessment of Artificial Intelligence Methods\nfor Prediction of hERG Channel Inhibition in the \u201cBig Data\u201d\nEra"],"prefix":"10.1021","volume":"60","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-5980-8288","authenticated-orcid":true,"given":"Vishal\nB.","family":"Siramshetty","sequence":"first","affiliation":[{"name":"National Center for Advancing Translational Sciences (NCATS) , 9800 Medical Center Drive , , ,","place":["Rockville, Maryland, United States, 20850"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Dac-Trung","family":"Nguyen","sequence":"additional","affiliation":[{"name":"National Center for Advancing Translational Sciences (NCATS) , 9800 Medical Center Drive , , ,","place":["Rockville, Maryland, United States, 20850"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Natalia J.","family":"Martinez","sequence":"additional","affiliation":[{"name":"National Center for Advancing Translational Sciences (NCATS) , 9800 Medical Center Drive , , ,","place":["Rockville, Maryland, United States, 20850"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Noel T.","family":"Southall","sequence":"additional","affiliation":[{"name":"National Center for Advancing Translational Sciences (NCATS) , 9800 Medical Center Drive , , ,","place":["Rockville, Maryland, United States, 20850"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Anton","family":"Simeonov","sequence":"additional","affiliation":[{"name":"National Center for Advancing Translational Sciences (NCATS) , 9800 Medical Center Drive , , ,","place":["Rockville, Maryland, United States, 20850"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2466-1711","authenticated-orcid":true,"given":"Alexey V.","family":"Zakharov","sequence":"additional","affiliation":[{"name":"National Center for Advancing Translational Sciences (NCATS) , 9800 Medical Center Drive , , ,","place":["Rockville, Maryland, United States, 20850"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"316","published-online":{"date-parts":[[2020,12,1]]},"reference":[{"key":"2026081804393791200_cit1","doi-asserted-by":"publisher","first-page":"833","DOI":"10.1038\/379833a0","article-title":"The inward rectification mechanism\nof the HERG cardiac potassium channel","volume":"379","author":"Smith","year":"1996","journal-title":"Nature"},{"key":"2026081804393791200_cit2","doi-asserted-by":"publisher","first-page":"1393","DOI":"10.1152\/physrev.00036.2011","article-title":"hERG K\u00ad(+) channels: structure, function, and clinical\nsignificance","volume":"92","author":"Vandenberg","year":"2012","journal-title":"Physiol. Rev."},{"key":"2026081804393791200_cit3","doi-asserted-by":"publisher","first-page":"299","DOI":"10.1016\/0092-8674(95)90340-2","article-title":"A mechanistic\nlink\nbetween an inherited and an acquired cardiac arrhythmia: HERG encodes\nthe IKr potassium channel","volume":"81","author":"Sanguinetti","year":"1995","journal-title":"Cell"},{"key":"2026081804393791200_cit4","doi-asserted-by":"publisher","first-page":"463","DOI":"10.1038\/nature04710","article-title":"hERG potassium channels and cardiac\narrhythmia","volume":"440","author":"Sanguinetti","year":"2006","journal-title":"Nature"},{"key":"2026081804393791200_cit5","doi-asserted-by":"publisher","first-page":"1216","DOI":"10.1053\/euhj.2000.2249","article-title":"The potential for QT prolongation\nand proarrhythmia by non-antiarrhythmic drugs: clinical and regulatory\nimplications. Report on a policy conference of the European Society\nof Cardiology","volume":"21","author":"Haverkamp","year":"2000","journal-title":"Eur. Heart J."},{"key":"2026081804393791200_cit6","doi-asserted-by":"publisher","first-page":"D1080","DOI":"10.1093\/nar\/gkv1192","article-title":"WITHDRAWN--a resource for withdrawn and discontinued\ndrugs","volume":"44","author":"Siramshetty","year":"2016","journal-title":"Nucleic Acids Res."},{"key":"2026081804393791200_cit7","doi-asserted-by":"publisher","first-page":"321","DOI":"10.1517\/14728222.11.3.321","article-title":"The hERG\npotassium channel as a therapeutic target","volume":"11","author":"Witchel","year":"2007","journal-title":"Expert\nOpin. Ther. Targets"},{"key":"2026081804393791200_cit8","doi-asserted-by":"publisher","first-page":"181","DOI":"10.1016\/j.phrs.2008.01.009","article-title":"The hERG K+ channel:\ntarget and antitarget strategies in drug development","volume":"57","author":"Raschi","year":"2008","journal-title":"Pharmacol. Res."},{"key":"2026081804393791200_cit9","doi-asserted-by":"publisher","first-page":"498","DOI":"10.1177\/0091270006286436","article-title":"Clinical\nevaluation of QT\/QTc prolongation\nand proarrhythmic potential for nonantiarrhythmic drugs: the International\nConference on Harmonization of Technical Requirements for Registration\nof Pharmaceuticals for Human Use E14 guideline","volume":"46","author":"Darpo","year":"2006","journal-title":"J. Clin. Pharmacol."},{"key":"2026081804393791200_cit10","doi-asserted-by":"publisher","first-page":"183","DOI":"10.1002\/jat.1395","article-title":"Collation, assessment and analysis\nof literature in vitro data on hERG receptor blocking potency for\nsubsequent modeling of drugs\u2019 cardiotoxic properties","volume":"29","author":"Polak","year":"2009","journal-title":"J. Appl. Toxicol."},{"key":"2026081804393791200_cit11","doi-asserted-by":"publisher","first-page":"278","DOI":"10.1080\/15376510902777194","article-title":"hERG in vitro interchange\nfactors--development and verification","volume":"19","author":"Wisniowska","year":"2009","journal-title":"Toxicol.\nMech. Methods"},{"key":"2026081804393791200_cit12","doi-asserted-by":"publisher","first-page":"65","DOI":"10.1016\/s1056-8719(03)00041-8","article-title":"Troubleshooting\nproblems with in vitro screening of drugs for QT interval prolongation\nusing HERG K+ channels expressed in mammalian cell lines and Xenopus\noocytes","volume":"48","author":"Witchel","year":"2002","journal-title":"J. Pharmacol. Toxicol. Methods"},{"key":"2026081804393791200_cit13","doi-asserted-by":"publisher","first-page":"1813","DOI":"10.1126\/science.1096361","article-title":"The many\nroles of computation in drug discovery","volume":"303","author":"Jorgensen","year":"2004","journal-title":"Science"},{"key":"2026081804393791200_cit14","doi-asserted-by":"publisher","first-page":"328","DOI":"10.2174\/138620712799361825","article-title":"Computer-aided drug\ndesign: lead discovery and optimization","volume":"15","author":"Xiang","year":"2012","journal-title":"Comb.\nChem. High Throughput Screen."},{"key":"2026081804393791200_cit15","doi-asserted-by":"publisher","first-page":"334","DOI":"10.1124\/pr.112.007336","article-title":"Computational\nmethods in drug discovery","volume":"66","author":"Sliwoski","year":"2014","journal-title":"Pharmacol.\nRev."},{"key":"2026081804393791200_cit16","doi-asserted-by":"publisher","first-page":"1399","DOI":"10.2174\/1568026614666140506124442","article-title":"Tuning HERG\nout: antitarget QSAR\nmodels for drug development","volume":"14","author":"Braga","year":"2014","journal-title":"Curr. Top. Med.\nChem."},{"key":"2026081804393791200_cit17","doi-asserted-by":"publisher","first-page":"1214","DOI":"10.1021\/acs.jcim.8b00124","article-title":"Multi-Descriptor\nRead Across (MuDRA): A Simple and\nTransparent Approach for Developing Accurate Quantitative Structure-Activity\nRelationship Models","volume":"58","author":"Alves","year":"2018","journal-title":"J. Chem. Inf. Model."},{"key":"2026081804393791200_cit18","doi-asserted-by":"publisher","first-page":"7","DOI":"10.3389\/fchem.2017.00007","article-title":"Computational Tool for Fast in silico\nEvaluation of\nhERG K+ Channel Affinity","volume":"5","author":"Chemi","year":"2017","journal-title":"Front. Chem."},{"key":"2026081804393791200_cit19","doi-asserted-by":"publisher","first-page":"229","DOI":"10.1007\/s10822-016-9898-z","article-title":"A k-nearest neighbor\nclassification of hERG K\u00ad(+) channel blockers","volume":"30","author":"Chavan","year":"2016","journal-title":"J. Comput. Aided Mol. Des."},{"key":"2026081804393791200_cit20","doi-asserted-by":"publisher","first-page":"2855","DOI":"10.1021\/acs.molpharmaceut.6b00471","article-title":"ADMET Evaluation\nin Drug Discovery. 16. Predicting hERG Blockers by Combining Multiple\nPharmacophores and Machine Learning Approaches","volume":"13","author":"Wang","year":"2016","journal-title":"Mol. Pharm."},{"key":"2026081804393791200_cit21","doi-asserted-by":"publisher","first-page":"1600126","DOI":"10.1002\/minf.201600126","article-title":"Prediction of hERG\nLiability - Using SVM Classification, Bootstrapping\nand Jackknifing","volume":"36","author":"Sun","year":"2017","journal-title":"Mol. Inf."},{"key":"2026081804393791200_cit22","doi-asserted-by":"publisher","first-page":"1224","DOI":"10.1021\/acs.jcim.8b00150","article-title":"The Catch-22\nof Predicting hERG Blockade Using Publicly Accessible Bioactivity\nData","volume":"58","author":"Siramshetty","year":"2018","journal-title":"J. Chem. Inf. Model."},{"key":"2026081804393791200_cit23","doi-asserted-by":"publisher","first-page":"12220","DOI":"10.1038\/s41598-019-47536-3","article-title":"Support Vector\nMachine\nmodel for hERG inhibitory activities based on the integrated hERG\ndatabase using descriptor selection by NSGA-II","volume":"9","author":"Ogura","year":"2019","journal-title":"Sci. Rep."},{"key":"2026081804393791200_cit24","doi-asserted-by":"publisher","first-page":"2615","DOI":"10.3390\/molecules25112615","article-title":"The Study\non the hERG Blocker Prediction\nUsing Chemical Fingerprint Analysis","volume":"25","author":"Choi","year":"2020","journal-title":"Molecules"},{"key":"2026081804393791200_cit25","doi-asserted-by":"publisher","first-page":"100089","DOI":"10.1016\/j.comtox.2019.100089","article-title":"hERG liability classification\nmodels using machine learning techniques","volume":"12","author":"Konda","year":"2019","journal-title":"Comput.\nToxicol."},{"key":"2026081804393791200_cit26","doi-asserted-by":"publisher","first-page":"263","DOI":"10.1021\/ci500747n","article-title":"Deep neural\nnets as a method for quantitative structure-activity relationships","volume":"55","author":"Ma","year":"2015","journal-title":"J. Chem. Inf. Model."},{"key":"2026081804393791200_cit27","doi-asserted-by":"publisher","first-page":"4462","DOI":"10.1021\/acs.molpharmaceut.7b00578","article-title":"Comparison of Deep\nLearning With Multiple Machine Learning Methods and Metrics Using\nDiverse Drug Discovery Data Sets","volume":"14","author":"Korotcov","year":"2017","journal-title":"Mol. Pharm."},{"key":"2026081804393791200_cit28","doi-asserted-by":"publisher","first-page":"497","DOI":"10.1186\/s12859-017-1895-2","article-title":"Development of models\nfor predicting Torsade de Pointes cardiac arrhythmias using perceptron\nneural networks","volume":"18","author":"Sharifi","year":"2017","journal-title":"BMC Bioinf."},{"key":"2026081804393791200_cit29","doi-asserted-by":"publisher","first-page":"1073","DOI":"10.1021\/acs.jcim.8b00769","article-title":"Deep Learning-Based Prediction of\nDrug-Induced Cardiotoxicity","volume":"59","author":"Cai","year":"2019","journal-title":"J. Chem. Inf. Model."},{"key":"2026081804393791200_cit30","doi-asserted-by":"publisher","first-page":"250","DOI":"10.1186\/s12859-019-2814-5","article-title":"Computational\ndetermination\nof hERG-related cardiotoxicity of drug candidates","volume":"20","author":"Lee","year":"2019","journal-title":"BMC Bioinf."},{"key":"2026081804393791200_cit31","doi-asserted-by":"publisher","first-page":"899","DOI":"10.1016\/j.tips.2017.06.004","article-title":"Towards\na Structural View of Drug\nBinding to hERG K\u00ad(+) Channels","volume":"38","author":"Vandenberg","year":"2017","journal-title":"Trends Pharmacol.\nSci."},{"key":"2026081804393791200_cit32","doi-asserted-by":"publisher","first-page":"513","DOI":"10.1039\/c7sc02664a","article-title":"MoleculeNet: a benchmark\nfor molecular machine learning","volume":"9","author":"Wu","year":"2018","journal-title":"Chem. Sci."},{"key":"2026081804393791200_cit33","doi-asserted-by":"publisher","first-page":"3370","DOI":"10.1021\/acs.jcim.9b01076","article-title":"Analyzing Learned Molecular Representations for Property\nPrediction","volume":"59","author":"Yang","year":"2019","journal-title":"J. Chem. Inf. Model."},{"key":"2026081804393791200_cit34","doi-asserted-by":"publisher","first-page":"131","DOI":"10.3390\/biom8040131","article-title":"Improving Chemical Autoencoder Latent\nSpace and Molecular\nDe Novo Generation Diversity with Heteroencoders","volume":"8","author":"Bjerrum","year":"2018","journal-title":"Biomolecules"},{"key":"2026081804393791200_cit35","doi-asserted-by":"publisher","first-page":"17","DOI":"10.3389\/frai.2019.00017","article-title":"Descriptor Free QSAR Modeling Using\nDeep Learning With\nLong Short-Term Memory Neural Networks","volume":"2","author":"Chakravarti","year":"2019","journal-title":"Front.\nArtif. Intell."},{"key":"2026081804393791200_cit36","doi-asserted-by":"publisher","first-page":"D1100","DOI":"10.1093\/nar\/gkr777","article-title":"ChEMBL:\na large-scale bioactivity database for drug discovery","volume":"40","author":"Gaulton","year":"2012","journal-title":"Nucleic Acids Res."},{"key":"2026081804393791200_cit37","doi-asserted-by":"publisher","first-page":"1243","DOI":"10.1021\/acs.jcim.6b00129","article-title":"Trust, but\nVerify II: A Practical\nGuide to Chemogenomics Data Curation","volume":"56","author":"Fourches","year":"2016","journal-title":"J. Chem.\nInf. Model."},{"key":"2026081804393791200_cit38","doi-asserted-by":"publisher","first-page":"2240","DOI":"10.1021\/ci400308z","article-title":"hERG Me Out","volume":"53","author":"Czodrowski","year":"2013","journal-title":"J. Chem. Inf. Model."},{"key":"2026081804393791200_cit39","doi-asserted-by":"publisher","first-page":"30","DOI":"10.1016\/j.ab.2009.07.003","article-title":"A new homogeneous\nhigh-throughput\nscreening assay for profiling compound activity on the human ether-a-go-go-related\ngene channel","volume":"394","author":"Titus","year":"2009","journal-title":"Anal. Biochem."},{"key":"2026081804393791200_cit40","unstructured":"Drugs@FDA: FDA-Approved Drugs. https:\/\/www.accessdata.fda.gov\/scripts\/cder\/daf\/index.cfm (accessed December 27, 2019)."},{"key":"2026081804393791200_cit41","doi-asserted-by":"publisher","first-page":"341","DOI":"10.1002\/phar.2087","article-title":"Validation and Clinical Utility of\nthe hERG IC50:Cmax Ratio to Determine the Risk of Drug-Induced Torsades\nde Pointes: A Meta-Analysis","volume":"38","author":"Lehmann","year":"2018","journal-title":"Pharmacotherapy"},{"key":"2026081804393791200_cit42","doi-asserted-by":"publisher","first-page":"1291","DOI":"10.1016\/j.drudis.2016.06.013","article-title":"Descriptors and\ntheir selection methods in QSAR analysis: paradigm for drug design","volume":"21","author":"Danishuddin","year":"2016","journal-title":"Drug Discovery Today"},{"key":"2026081804393791200_cit43","doi-asserted-by":"publisher","first-page":"1538","DOI":"10.1016\/j.drudis.2018.05.010","article-title":"Machine learning\nin chemoinformatics and drug discovery","volume":"23","author":"Lo","year":"2018","journal-title":"Drug\nDiscovery Today"},{"key":"2026081804393791200_cit44","unstructured":"RDKit: Open-Source Cheminformatics Software. https:\/\/www.rdkit.org\/ (accessed\nDecember 1, 2019)."},{"key":"2026081804393791200_cit45","doi-asserted-by":"publisher","first-page":"268","DOI":"10.1021\/acscentsci.7b00572","article-title":"Automatic\nChemical Design Using a Data-Driven Continuous Representation of Molecules","volume":"4","author":"Go\u0301mez-Bombarelli","year":"2018","journal-title":"ACS Cent. Sci."},{"key":"2026081804393791200_cit46","doi-asserted-by":"publisher","first-page":"1692","DOI":"10.1039\/c8sc04175j","article-title":"Learning continuous\nand data-driven molecular descriptors by translating equivalent chemical\nrepresentations","volume":"10","author":"Winter","year":"2019","journal-title":"Chem. Sci."},{"key":"2026081804393791200_cit47","doi-asserted-by":"publisher","first-page":"1735","DOI":"10.1162\/neco.1997.9.8.1735","article-title":"Long short-term memory","volume":"9","author":"Hochreiter","year":"1997","journal-title":"Neural\nComput."},{"key":"2026081804393791200_cit48","doi-asserted-by":"publisher","first-page":"2222","DOI":"10.1109\/tnnls.2016.2582924","article-title":"LSTM: A Search\nSpace Odyssey","volume":"28","author":"Greff","year":"2017","journal-title":"IEEE Trans. Neural. Netw. Learn\nSyst."},{"key":"2026081804393791200_cit49","doi-asserted-by":"publisher","first-page":"71","DOI":"10.1186\/s13321-019-0393-0","article-title":"Randomized SMILES strings\nimprove the quality of molecular generative models","volume":"11","author":"Aru\u0301s-Pous","year":"2019","journal-title":"J. Cheminf."},{"key":"2026081804393791200_cit50","doi-asserted-by":"publisher","DOI":"10.26434\/chemrxiv.7097960","article-title":"DeepSMILES: An Adaptation\nof SMILES for Use in Machine-Learning of\nChemical Structures","author":"O\u2019Boyle","year":"2018","journal-title":"ChemRxiv"},{"key":"2026081804393791200_cit51","doi-asserted-by":"publisher","first-page":"120","DOI":"10.1021\/acscentsci.7b00512","article-title":"Generating Focused\nMolecule Libraries for Drug Discovery with Recurrent Neural Networks","volume":"4","author":"Segler","year":"2018","journal-title":"ACS Cent. Sci."},{"key":"2026081804393791200_cit52","doi-asserted-by":"publisher","first-page":"6728","DOI":"10.1016\/j.bmcl.2011.09.051","article-title":"One-class\nclassification as a novel method of ligand-based virtual screening:\nthe case of glycogen synthase kinase 3beta inhibitors","volume":"21","author":"Karpov","year":"2011","journal-title":"Bioorg. Med. Chem. Lett."},{"key":"2026081804393791200_cit53","doi-asserted-by":"publisher","first-page":"1096","DOI":"10.1021\/acs.jcim.8b00839","article-title":"GuacaMol:\nBenchmarking\nModels for de Novo Molecular Design","volume":"59","author":"Brown","year":"2019","journal-title":"J. Chem.\nInf. Model."},{"key":"2026081804393791200_cit54","unstructured":"scikit-learn: Machine Learning in Python. https:\/\/scikit-learn.org\/stable\/ (accessed December 4, 2019)."},{"key":"2026081804393791200_cit55","unstructured":"Keras:\nThe Python Deep Learning library. https:\/\/keras.io\/ (accessed December\n4, 2019)."},{"key":"2026081804393791200_cit56","unstructured":"TensorFlow. https:\/\/www.tensorflow.org\/ (accessed December 4, 2019)."},{"key":"2026081804393791200_cit57","unstructured":"DeepChem. https:\/\/deepchem.io\/ (accessed\nDecember 4, 2019)."},{"key":"2026081804393791200_cit58","doi-asserted-by":"publisher","first-page":"832","DOI":"10.1109\/34.709601","article-title":"The random\nsubspace method for constructing decision forests","volume":"20","author":"Ho","year":"1998","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"2026081804393791200_cit59","doi-asserted-by":"publisher","first-page":"600","DOI":"10.3389\/fgene.2019.00600","article-title":"Five-Feature Model\nfor Developing the Classifier for Synergistic vs. Antagonistic Drug\nCombinations Built by XGBoost","volume":"10","author":"Ji","year":"2019","journal-title":"Front Genet."},{"key":"2026081804393791200_cit60","doi-asserted-by":"publisher","first-page":"450","DOI":"10.1021\/ci970100x","article-title":"Aqueous solubility\nprediction of\ndrugs based on molecular topology and neural network modeling","volume":"38","author":"Huuskonen","year":"1998","journal-title":"J. Chem. Inf. Comput. Sci."},{"key":"2026081804393791200_cit61","doi-asserted-by":"publisher","first-page":"903","DOI":"10.1021\/ci0203702","article-title":"On the use of neural network ensembles\nin QSAR and QSPR","volume":"42","author":"Agrafiotis","year":"2002","journal-title":"J. Chem. Inf. Comput. Sci."},{"key":"2026081804393791200_cit62","doi-asserted-by":"publisher","first-page":"1460","DOI":"10.1021\/ci020039i","article-title":"QSAR\/QSPR studies using probabilistic neural networks\nand generalized regression neural networks","volume":"42","author":"Mosier","year":"2002","journal-title":"J. Chem. Inf. Comput. Sci."},{"key":"2026081804393791200_cit63","doi-asserted-by":"publisher","first-page":"139","DOI":"10.1385\/mb:27:2:139","article-title":"Neural\nnetworks as robust tools in drug lead discovery and development","volume":"27","author":"Winkler","year":"2004","journal-title":"Mol. Biotechnol."},{"key":"2026081804393791200_cit64","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1016\/s0167-7012(00)00201-3","article-title":"Artificial neural networks:\nfundamentals, computing,\ndesign, and application","volume":"43","author":"Basheer","year":"2000","journal-title":"J. Microbiol. Methods"},{"key":"2026081804393791200_cit65","unstructured":"Bahdanau, D.; Cho, K.; Bengio, Y.\n          Neural\nMachine Translation\nby Jointly Learning to Align and Translate. International Conference on Learning Representations, 2015, 2015; Vol. abs\/1409.0473."},{"key":"2026081804393791200_cit66","doi-asserted-by":"crossref","unstructured":"Luong, T.; Pham, H.; Manning, C. D.\n          Effective Approaches\nto Attention-based\nNeural Machine Translation. Empirical Methods\nin Natural Language Processing, 2015, 2015;\nVol. 10.18653\/v1\/D15-1.","DOI":"10.18653\/v1\/D15-1166"},{"key":"2026081804393791200_cit67","unstructured":"Keras-Self-Attention. https:\/\/github.com\/CyberZHG\/keras-self-attention (accessed December 4, 2019)."},{"key":"2026081804393791200_cit68","unstructured":"StarDrop. https:\/\/www.optibrium.com\/stardrop\/ (accessed December 17, 2019)."},{"key":"2026081804393791200_cit69","doi-asserted-by":"publisher","first-page":"698","DOI":"10.1002\/minf.201500040","article-title":"Pred-hERG: A Novel\nweb-Accessible Computational Tool for Predicting Cardiac Toxicity","volume":"34","author":"Braga","year":"2015","journal-title":"Mol. Inf."},{"key":"2026081804393791200_cit70","unstructured":"Pred-hERG 4.2. http:\/\/predherg.labmol.com.br\/ (accessed December 17, 2019)."},{"key":"2026081804393791200_cit71","doi-asserted-by":"publisher","first-page":"1947","DOI":"10.1021\/ci034160g","article-title":"Random forest: a classification and\nregression tool for compound classification and QSAR modeling","volume":"43","author":"Svetnik","year":"2003","journal-title":"J. Chem. Inf. Comput. Sci."},{"key":"2026081804393791200_cit72","doi-asserted-by":"publisher","first-page":"878","DOI":"10.15252\/msb.20156651","article-title":"Deep learning for computational\nbiology","volume":"12","author":"Angermueller","year":"2016","journal-title":"Mol. Syst. Biol."},{"key":"2026081804393791200_cit73","doi-asserted-by":"publisher","first-page":"1253","DOI":"10.1021\/acs.jcim.8b00785","article-title":"Predictive Multitask Deep Neural\nNetwork Models for ADME-Tox Properties: Learning from Large Data Sets","volume":"59","author":"Wenzel","year":"2019","journal-title":"J. Chem. Inf. Model."},{"key":"2026081804393791200_cit74","doi-asserted-by":"publisher","first-page":"1445","DOI":"10.1021\/acs.molpharmaceut.5b00982","article-title":"Applications\nof Deep Learning in Biomedicine","volume":"13","author":"Mamoshina","year":"2016","journal-title":"Mol. Pharm."},{"key":"2026081804393791200_cit75","doi-asserted-by":"publisher","first-page":"638","DOI":"10.21105\/joss.00638","article-title":"MLxtend: Providing\nmachine learning and data science utilities and extensions to Python\u2019s\nscientific computing stack","volume":"3","author":"Raschka","year":"2018","journal-title":"J. Open Source Software"},{"key":"2026081804393791200_cit76","doi-asserted-by":"publisher","first-page":"177","DOI":"10.1021\/ci049714+","article-title":"ZINC--a free database of commercially available compounds\nfor virtual screening","volume":"45","author":"Irwin","year":"2005","journal-title":"J. Chem. Inf. Model."},{"key":"2026081804393791200_cit77","article-title":"SMILES Enumeration\nas Data Augmentation for Neural Network Modeling of Molecules","author":"Bjerrum","year":"2017"},{"key":"2026081804393791200_cit78","doi-asserted-by":"publisher","first-page":"120","DOI":"10.1021\/acscentsci.7b00512","article-title":"Generating Focused\nMolecule Libraries for Drug Discovery with Recurrent Neural Networks","volume":"4","author":"Segler","year":"2018","journal-title":"ACS Cent. Sci."},{"key":"2026081804393791200_cit79","doi-asserted-by":"publisher","first-page":"48","DOI":"10.1186\/s13321-017-0235-x","article-title":"Molecular de-novo design\nthrough deep reinforcement learning","volume":"9","author":"Olivecrona","year":"2017","journal-title":"J. Cheminf."},{"key":"2026081804393791200_cit80","doi-asserted-by":"publisher","first-page":"1700111","DOI":"10.1002\/minf.201880141","article-title":"Generative Recurrent Networks for\nDe Novo Drug Design","volume":"37","author":"Gupta","year":"2018","journal-title":"Mol. Inf."},{"key":"2026081804393791200_cit81","doi-asserted-by":"publisher","DOI":"10.1126\/sciadv.aap7885","article-title":"Deep reinforcement learning for de\nnovo drug design","volume":"4","author":"Popova","year":"2018","journal-title":"Sci. Adv."},{"key":"2026081804393791200_cit82","doi-asserted-by":"publisher","first-page":"1182","DOI":"10.1021\/acs.jcim.8b00751","article-title":"De Novo Molecular\nDesign by Combining Deep Autoencoder Recurrent Neural Networks with\nGenerative Topographic Mapping","volume":"59","author":"Sattarov","year":"2019","journal-title":"J. Chem. Inf.\nModel."},{"key":"2026081804393791200_cit83","doi-asserted-by":"publisher","first-page":"1038","DOI":"10.1038\/s41587-019-0224-x","article-title":"Deep learning enables rapid identification of potent\nDDR1 kinase inhibitors","volume":"37","author":"Zhavoronkov","year":"2019","journal-title":"Nat. Biotechnol."},{"key":"2026081804393791200_cit84","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.0c00174","article-title":"The Synthesizability\nof Molecules Proposed by Generative\nModels","author":"Gao","year":"2020","journal-title":"J. Chem. Inf. Model."}],"container-title":["Journal of Chemical Information and Modeling"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/pubs.acs.org\/doi\/pdf\/10.1021\/acs.jcim.0c00884","content-type":"application\/pdf","content-version":"vor","intended-application":"unspecified"},{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article-pdf\/60\/12\/6007\/9623810\/ci0c00884.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article-pdf\/60\/12\/6007\/9623810\/ci0c00884.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,18]],"date-time":"2026-08-18T09:27:57Z","timestamp":1787045277000},"score":1,"resource":{"primary":{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article\/60\/12\/6007\/891333\/Critical-Assessment-of-Artificial-Intelligence"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2020,12,1]]},"references-count":84,"journal-issue":{"issue":"12","published-online":{"date-parts":[[2020,12,1]]},"published-print":{"date-parts":[[2020,12,28]]}},"URL":"https:\/\/doi.org\/10.1021\/acs.jcim.0c00884","relation":{"has-preprint":[{"id-type":"doi","id":"10.26434\/chemrxiv.12119040","asserted-by":"object"},{"id-type":"doi","id":"10.26434\/chemrxiv.12119040.v1","asserted-by":"object"}]},"ISSN":["1549-9596","1549-960X"],"issn-type":[{"value":"1549-9596","type":"print"},{"value":"1549-960X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2020,12,1]]}}}