{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,10]],"date-time":"2026-08-10T13:28:21Z","timestamp":1786368501700,"version":"build-2736575974"},"reference-count":46,"publisher":"American Chemical Society (ACS)","issue":"15","license":[{"start":{"date-parts":[[2026,7,28]],"date-time":"2026-07-28T00:00:00Z","timestamp":1785196800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,7,28]],"date-time":"2026-07-28T00:00:00Z","timestamp":1785196800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"},{"start":{"date-parts":[[2026,7,28]],"date-time":"2026-07-28T00:00:00Z","timestamp":1785196800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-045"}],"funder":[{"DOI":"10.13039\/100004334","name":"Merck","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100004334","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,8,10]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Chemical analog design in the hit-to-lead and lead optimization stages of drug discovery relies on systematic structural modifications, often guided by medicinal chemistry intuition. Although a matched molecular pair (MMP) provides an interpretable framework to capture intuition, models trained on individual MMP instances face significant limitations, such as bias toward frequent transformations in historical data. This work introduces a novel approach to transform the concept of \u201chow to design chemical analogs like a medicinal chemist\u201d into a primary training objective for generative models by focusing on matched molecular pair transformations (MMPTs) as the fundamental unit of chemical modifications. This allows for a more generalizable and context-independent representation of medicinal chemistry intuition, enabling the application of the same transformation priors across different projects, regardless of the target or indication. Using a consistently curated ChEMBL-derived data set, we compared a transformation-centric foundation model (MMPT-FM) with multiple MMP-based generative formulations trained on the same underlying data. Furthermore, performance is assessed through challenging within-patent and cross-patent real-world test cases derived from drug discovery patents. The MMPT-FM model achieves comparable or improved recall metrics across all test cases, demonstrating particularly strong performance for low-frequency and previously unseen transformations. This work not only shifts the paradigm of learning and utilizes medicinal chemistry intuition in an efficient and scalable manner in the AI for drug discovery era but also establishes a significant competitive advantage by enabling the drug discovery industry to encode decades of collective medicinal chemistry knowledge\u2500both public and proprietary\u2500into a scalable foundation generative model that could help researchers design chemical analogs.<\/jats:p>","DOI":"10.1021\/acs.jcim.6c01032","type":"journal-article","created":{"date-parts":[[2026,8,4]],"date-time":"2026-08-04T13:13:55Z","timestamp":1785849235000},"page":"8908-8922","source":"Crossref","is-referenced-by-count":0,"title":["Scalable and Generalizable\nAnalog Design via Learning\nMedicinal Chemistry Intuition from Matched Molecular Pair Transformations"],"prefix":"10.1021","volume":"66","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-9381-7500","authenticated-orcid":true,"given":"Hao-Wei","family":"Pang","sequence":"first","affiliation":[{"name":"Merck & Co., Inc. , , ,","place":["Rahway, New Jersey, United States, 07065"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9099-4357","authenticated-orcid":true,"given":"Peter Zhiping","family":"Zhang","sequence":"additional","affiliation":[{"name":"Merck & Co., Inc. , , ,","place":["Rahway, New Jersey, United States, 07065"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-7501-7581","authenticated-orcid":true,"given":"Bo","family":"Pan","sequence":"additional","affiliation":[{"name":"Emory University , , , ,","place":["Atlanta, Georgia, United States, 30322"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2648-9989","authenticated-orcid":true,"given":"Liang","family":"Zhao","sequence":"additional","affiliation":[{"name":"Emory University , , , ,","place":["Atlanta, Georgia, United States, 30322"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8217-5896","authenticated-orcid":true,"given":"Xiang","family":"Yu","sequence":"additional","affiliation":[{"name":"Merck & Co., Inc. , , ,","place":["Rahway, New Jersey, United States, 07065"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Liying","family":"Zhang","sequence":"additional","affiliation":[{"name":"Merck & Co., Inc. , , ,","place":["Rahway, New Jersey, United States, 07065"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"316","published-online":{"date-parts":[[2026,7,28]]},"reference":[{"key":"2026081008092793600_cit1","first-page":"12","article-title":"Analogue-based drug discovery","volume":"32","author":"Fischer","year":"2010","journal-title":"Chemistry International\u2013Newsmagazine For IUPAC"},{"key":"2026081008092793600_cit2","doi-asserted-by":"publisher","first-page":"3212","DOI":"10.1021\/jm900107b","article-title":"Exploration of Structure-\nActivity Relationship Determinants in Analogue Series","volume":"52","author":"Peltason","year":"2009","journal-title":"J. Med. Chem."},{"key":"2026081008092793600_cit3","doi-asserted-by":"publisher","first-page":"6651","DOI":"10.1038\/s41467-023-42242-1","article-title":"Extracting\nmedicinal chemistry intuition via preference machine learning","volume":"14","author":"Choung","year":"2023","journal-title":"Nat. Commun."},{"key":"2026081008092793600_cit4","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1016\/S0169-409X(96)00423-1","article-title":"Experimental and\ncomputational approaches to estimate solubility and permeability in\ndrug discovery and development settings","volume":"23","author":"Lipinski","year":"1997","journal-title":"Adv.\nDrug Delivery Rev."},{"key":"2026081008092793600_cit5","doi-asserted-by":"publisher","first-page":"2719","DOI":"10.1021\/jm901137j","article-title":"New substructure filters for removal\nof pan assay interference\ncompounds (PAINS) from screening libraries and for their exclusion\nin bioassays","volume":"53","author":"Baell","year":"2010","journal-title":"J. Med. Chem."},{"key":"2026081008092793600_cit6","doi-asserted-by":"publisher","first-page":"10860","DOI":"10.1021\/acs.jmedchem.4c03066","article-title":"Harnessing\nMedicinal\nChemical Intuition from Collective Intelligence","volume":"68","author":"Llompart","year":"2025","journal-title":"J. Med. Chem."},{"key":"2026081008092793600_cit7","doi-asserted-by":"publisher","first-page":"7823","DOI":"10.3390\/ijms21217823","article-title":"Discovery and Evaluation of Enantiopure\n9 H-pyrimido [4, 5-b] indoles as Nanomolar GSK-3\u03b2 Inhibitors with Improved Metabolic Stability","volume":"21","author":"Andreev","year":"2020","journal-title":"Int. J. Mol. Sci."},{"key":"2026081008092793600_cit8","first-page":"245","article-title":"Synopsis of Some\nRecent Tactical Application of Bioisosteres in Drug Design","volume-title":"Advances in Heterocyclic Chemistry","author":"Meanwell","year":"2017"},{"key":"2026081008092793600_cit9","doi-asserted-by":"publisher","first-page":"12290","DOI":"10.1021\/acs.jmedchem.0c00530","article-title":"Amide bond bioisosteres:\nStrategies, synthesis, and successes","volume":"63","author":"Kumari","year":"2020","journal-title":"J. Med.\nChem."},{"key":"2026081008092793600_cit10","doi-asserted-by":"publisher","first-page":"1872","DOI":"10.1021\/ci100258p","article-title":"Lead optimization using matched molecular pairs: inclusion\nof contextual information for enhanced prediction of HERG inhibition,\nsolubility, and lipophilicity","volume":"50","author":"Papadatos","year":"2010","journal-title":"J. Chem. Inf.\nModel"},{"key":"2026081008092793600_cit11","doi-asserted-by":"publisher","first-page":"7739","DOI":"10.1021\/jm200452d","article-title":"Matched molecular\npairs as a medicinal chemistry tool: miniperspective","volume":"54","author":"Griffen","year":"2011","journal-title":"J. Med. Chem."},{"key":"2026081008092793600_cit12","doi-asserted-by":"publisher","first-page":"902","DOI":"10.1021\/acs.jcim.8b00173","article-title":"mmpdb: An open-source matched molecular\npair platform for large multiproperty data sets","volume":"58","author":"Dalke","year":"2018","journal-title":"J. Chem. Inf. Model"},{"key":"2026081008092793600_cit13","first-page":"2323","article-title":"Junction tree\nvariational autoencoder for molecular graph generation","volume-title":"International conference on machine learning","author":"Jin","year":"2018"},{"key":"2026081008092793600_cit14","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":"2026081008092793600_cit15","doi-asserted-by":"publisher","first-page":"1734","DOI":"10.1021\/acs.jcim.2c01618","article-title":"Can We\nQuickly Learn to \u201cTranslate\u201d\nBioactive Molecules with Transformer Models?","volume":"63","author":"Tysinger","year":"2023","journal-title":"J. Chem. Inf. Model"},{"key":"2026081008092793600_cit16","doi-asserted-by":"publisher","first-page":"7315","DOI":"10.1038\/s41467-024-51672-4","article-title":"Exhaustive local chemical\nspace exploration using a transformer model","volume":"15","author":"Tibo","year":"2024","journal-title":"Nat. Commun"},{"key":"2026081008092793600_cit17","doi-asserted-by":"publisher","first-page":"18","DOI":"10.1186\/s13321-022-00599-3","article-title":"Transformer-based molecular\noptimization beyond matched molecular pairs","volume":"14","author":"He","year":"2022","journal-title":"J. Cheminform"},{"key":"2026081008092793600_cit18","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/s13321-021-00497-0","article-title":"Molecular optimization by capturing chemist\u2019s\nintuition using deep neural networks","volume":"13","author":"He","year":"2021","journal-title":"J. Cheminform"},{"key":"2026081008092793600_cit19","doi-asserted-by":"publisher","first-page":"20","DOI":"10.1186\/s13321-024-00812-5","article-title":"Reinvent 4:\nModern AI\u2013driven generative molecule\ndesign","volume":"16","author":"Loeffler","year":"2024","journal-title":"J. Cheminform"},{"key":"2026081008092793600_cit20","doi-asserted-by":"publisher","first-page":"1359","DOI":"10.1038\/s42256-024-00916-5","article-title":"Leveraging language\nmodel for advanced multiproperty\nmolecular optimization via prompt engineering","volume":"11","author":"Wu","year":"2024","journal-title":"Nat. Mach. Intell"},{"key":"2026081008092793600_cit21","doi-asserted-by":"publisher","first-page":"3609","DOI":"10.1021\/acs.jmedchem.5c03746","article-title":"Integrating Medicinal Chemist Expertise with\nDeep Learning for Automated\nMolecular Optimization","volume":"69","author":"Liang","year":"2026","journal-title":"J. Med. Chem"},{"key":"2026081008092793600_cit22","article-title":"Retrieval-Augmented Foundation Models for Matched Molecular Pair\nTransformations to Recapitulate Medicinal Chemistry Intuition","author":"Pan","year":"2026","journal-title":"arXiv"},{"key":"2026081008092793600_cit23","doi-asserted-by":"publisher","first-page":"D1100","DOI":"10.1093\/nar\/gkr777","article-title":"ChEMBL: a large-scale\nbioactivity database for drug discovery","volume":"40","author":"Gaulton","year":"2012","journal-title":"Nucleic\nAcids Res."},{"key":"2026081008092793600_cit24","unstructured":"Noutahi, E.; Mary, H.; Kovary, K. M.; Whitfield, S.; St-Laurent, J.; Hounwanou, H.; Craig, M.\n          datamol-io\/medchem: molecular filtering\nfor drug discovery (v2.0.5-alpha). Zenodo, 2025; 10.5281\/zenodo.14588938."},{"key":"2026081008092793600_cit25","unstructured":"Walters, P.\n          \n          rd_filters. 2018; https:\/\/github.com\/PatWalters\/rd_filters\/blob\/master\/rd_filters\/data\/alert_collection.csv."},{"key":"2026081008092793600_cit26","doi-asserted-by":"publisher","first-page":"2046","DOI":"10.1021\/acs.jcim.1c00469","article-title":"LibINVENT:\nreaction-based generative\nscaffold decoration for in silico library design","volume":"62","author":"Fialkova\u0301","year":"2022","journal-title":"J. Chem. Inf. Model"},{"key":"2026081008092793600_cit27","doi-asserted-by":"publisher","first-page":"392","DOI":"10.1039\/D2DD00115B","article-title":"Link-INVENT: generative linker design\nwith reinforcement learning","volume":"2","author":"Guo","year":"2023","journal-title":"Digital Discovery"},{"key":"2026081008092793600_cit28","first-page":"6140","article-title":"Exploring\nthe limits of transfer learning with a unified text-to-text transformer","volume":"21","author":"Raffel","year":"2020","journal-title":"J. Mach. Learn. Res."},{"key":"2026081008092793600_cit29","doi-asserted-by":"publisher","first-page":"270","DOI":"10.1162\/neco.1989.1.2.270","article-title":"A learning algorithm for continually running fully\nrecurrent neural networks","volume":"1","author":"Williams","year":"1989","journal-title":"Neural Comput."},{"key":"2026081008092793600_cit30","first-page":"6140","article-title":"Unifying molecular\nand textual representations via\nmulti-task language modelling","volume":"202","author":"Christofidellis","year":"2023","journal-title":"Int. Conf. Mach.\nLear"},{"key":"2026081008092793600_cit31","doi-asserted-by":"publisher","first-page":"28","DOI":"10.1186\/s13321-020-00431-w","article-title":"CReM: chemically\nreasonable mutations framework for structure generation","volume":"12","author":"Polishchuk","year":"2020","journal-title":"J. Cheminform."},{"key":"2026081008092793600_cit32","unstructured":"Vu, B.; Dominique, R.; Li, H.\n          Methods and Compounds\nfor Restoring\nMutant p53 Function. WO 2,017,143,291 A1, 2017."},{"key":"2026081008092793600_cit33","unstructured":"Vu, B.; Dominique, R.; Li, H.; Fahr, B.; Good, A.\n          Methods and Compounds for Restoring Mutant p53 Function. WO 2,021,061,643 A1, 2021."},{"key":"2026081008092793600_cit34","doi-asserted-by":"publisher","first-page":"8","DOI":"10.1186\/1758-2946-1-8","article-title":"Estimation of synthetic accessibility score of drug-like\nmolecules based on molecular complexity and fragment contributions","volume":"1","author":"Ertl","year":"2009","journal-title":"J. Cheminform."},{"key":"2026081008092793600_cit35","unstructured":"Landrum, G.\n          \n          RDKit: Open-source cheminformatics. 2016; http:\/\/www.rdkit.org."},{"key":"2026081008092793600_cit36","doi-asserted-by":"publisher","first-page":"70","DOI":"10.1186\/s13321-020-00472-1","article-title":"AiZynthFinder: a fast,\nrobust and flexible open-source\nsoftware for retrosynthetic planning","volume":"12","author":"Genheden","year":"2020","journal-title":"J. Cheminform"},{"key":"2026081008092793600_cit37","doi-asserted-by":"publisher","first-page":"1238","DOI":"10.1021\/acs.jmedchem.6b01437","article-title":"Recent advances in scaffold hopping:\nminiperspective","volume":"60","author":"Hu","year":"2017","journal-title":"J. Med. Chem."},{"key":"2026081008092793600_cit38","doi-asserted-by":"publisher","first-page":"103845","DOI":"10.1016\/j.drudis.2023.103845","article-title":"Molecular medicinal insights into\nscaffold hopping-based\ndrug discovery success","volume":"29","author":"Acharya","year":"2024","journal-title":"Drug Discovery Today"},{"key":"2026081008092793600_cit39","first-page":"283","volume-title":"Tactics in Contemporary\nDrug Design","author":"Meanwell","year":"2014"},{"key":"2026081008092793600_cit40","doi-asserted-by":"publisher","first-page":"6337","DOI":"10.1021\/acs.jmedchem.5c03171","article-title":"What Happens\nin Successful Optimizations? A Survey of 2018\u20132024 Literature","volume":"69","author":"Leeson","year":"2026","journal-title":"J. Med. Chem"},{"key":"2026081008092793600_cit41","doi-asserted-by":"publisher","first-page":"124","DOI":"10.1080\/00498250601050412","article-title":"Attenuating pregnane\nX receptor (PXR) activation: a\nmolecular modelling approach","volume":"37","author":"Gao","year":"2007","journal-title":"Xenobiotica"},{"key":"2026081008092793600_cit42","doi-asserted-by":"publisher","first-page":"1396","DOI":"10.1021\/acsmedchemlett.1c00187","article-title":"Overcoming the pregnane\nX receptor liability: rational\ndesign to eliminate PXR-mediated CYP induction","volume":"12","author":"Ramanjulu","year":"2021","journal-title":"ACS Med. Chem. Lett."},{"key":"2026081008092793600_cit43","doi-asserted-by":"publisher","first-page":"14261","DOI":"10.1021\/acs.jmedchem.2c01147","article-title":"Hydrogen-bond\ndonors in drug design","volume":"65","author":"Kenny","year":"2022","journal-title":"J. Med. Chem."},{"key":"2026081008092793600_cit44","doi-asserted-by":"publisher","first-page":"1000","DOI":"10.1021\/ci034243x","article-title":"ESOL: estimating\naqueous solubility directly from molecular structure","volume":"44","author":"Delaney","year":"2004","journal-title":"J. Chem. Inf. Comput. Sci."},{"key":"2026081008092793600_cit45","first-page":"12565","article-title":"mCLM: A Modular Chemical\nLanguage Model that Generates\nFunctional and Makeable Molecules","author":"Edwards","year":"2025","journal-title":"arXiv"},{"key":"2026081008092793600_cit200","unstructured":"Qiao, Y.; Pan, B.; Pang, H. W.; Zhang, P. Z.; Zhang, L.; Zhao, L.\n          MolWorld: Molecule World Models\nfor Actionable Molecular Optimization.\n          arXiv preprint, 2026, arXiv:2605.08954."}],"container-title":["Journal of Chemical Information\nand Modeling"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/pubs.acs.org\/doi\/pdf\/10.1021\/acs.jcim.6c01032","content-type":"application\/pdf","content-version":"vor","intended-application":"unspecified"},{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article-pdf\/66\/15\/8908\/66318976\/acs.jcim.6c01032.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article-pdf\/66\/15\/8908\/66318976\/acs.jcim.6c01032.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,10]],"date-time":"2026-08-10T12:31:14Z","timestamp":1786365074000},"score":1,"resource":{"primary":{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article\/66\/15\/8908\/5236113\/Scalable-and-Generalizable-Analog-Design-via"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,28]]},"references-count":46,"journal-issue":{"issue":"15","published-online":{"date-parts":[[2026,7,28]]},"published-print":{"date-parts":[[2026,8,10]]}},"URL":"https:\/\/doi.org\/10.1021\/acs.jcim.6c01032","relation":{"has-preprint":[{"id-type":"doi","id":"10.26434\/chemrxiv.15001722\/v2","asserted-by":"object"}]},"ISSN":["1549-9596","1549-960X"],"issn-type":[{"value":"1549-9596","type":"print"},{"value":"1549-960X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,28]]}}}