{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,14]],"date-time":"2026-09-14T14:18:37Z","timestamp":1789395517152,"version":"build-2803163510"},"reference-count":49,"publisher":"American Chemical Society (ACS)","issue":"17","license":[{"start":{"date-parts":[[2026,8,18]],"date-time":"2026-08-18T00:00:00Z","timestamp":1787011200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,8,18]],"date-time":"2026-08-18T00:00:00Z","timestamp":1787011200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"},{"start":{"date-parts":[[2026,8,18]],"date-time":"2026-08-18T00:00:00Z","timestamp":1787011200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-045"}],"funder":[{"DOI":"10.13039\/100000060","name":"National Institute of Allergy and Infectious Diseases","doi-asserted-by":"publisher","award":["U19-AI171954"],"award-info":[{"award-number":["U19-AI171954"]}],"id":[{"id":"10.13039\/100000060","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,9,14]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Fragment-based drug discovery (FBDD) relies heavily on the design of chemically viable linkers to connect fragments binding to different pocket regions into potent lead molecules. While recent generative models have advanced spatial fragment linking, they frequently produce linkers characterized by high torsional strain and non-drug-like motifs. In this work, we present LinkLlama, a fine-tuned Meta Llama 3 model that bridges the gap between text-based generation and 3D spatial awareness. By accepting natural language prompts that specify geometric constraints, such as distances and angles, alongside physicochemical targets like Lipinski\u2019s rules and rotatable bond limits, LinkLlama generates highly tailored molecules for the input fragments. Leveraging the inherent chemical grammar captured through supervised fine-tuning on a curated corpus of drug-like molecules from ChEMBL, the model prioritizes chemical validity without requiring complex reinforcement learning loops. Benchmarking on the ZINC and HiQBind data sets demonstrates that LinkLlama maintains competitive geometric fidelity compared to strictly 3D-aware models while achieving a 2-fold increase in the proportion of chemically reasonable designs. This rising success rate, jumping from 35% to over 80%, is defined by strict adherence to comprehensive structural filters including PAINS, non-drug-like chemical patterns, and complex ring systems. We further illustrate the model\u2019s versatility through prospective case studies in novel small-molecule scaffold hopping and PROTAC linker design, validated via molecular docking and molecular dynamics simulations against known crystal poses. Ultimately, LinkLlama demonstrates that large language models can overcome the structural pitfalls of purely 3D-generative methods, offering a highly controllable and chemically robust framework to accelerate linker design and drug discovery in general.<\/jats:p>","DOI":"10.1021\/acs.jcim.6c01261","type":"journal-article","created":{"date-parts":[[2026,8,18]],"date-time":"2026-08-18T16:13:18Z","timestamp":1787069598000},"page":"11348-11360","source":"Crossref","is-referenced-by-count":0,"title":["LinkLlama: Enabling\na Large Language Model for Chemically\nReasonable Linker Design"],"prefix":"10.1021","volume":"66","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-6472-1665","authenticated-orcid":true,"given":"Kunyang","family":"Sun","sequence":"first","affiliation":[{"name":"University of California , , , ,","place":["Berkeley, California, United States, 94720"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1706-3791","authenticated-orcid":true,"given":"Yingze","family":"Wang","sequence":"additional","affiliation":[{"name":"University of California , , , ,","place":["Berkeley, California, United States, 94720"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Justin","family":"Purnomo","sequence":"additional","affiliation":[{"name":"University of California , , , ,","place":["Berkeley, California, United States, 94720"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Joseph M.","family":"Cavanagh","sequence":"additional","affiliation":[{"name":"University of California , , , ,","place":["Berkeley, California, United States, 94720"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Giovanni Battista","family":"Alteri","sequence":"additional","affiliation":[{"name":"University of California , , , ,","place":["Berkeley, California, United States, 94720"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0025-8987","authenticated-orcid":true,"given":"Teresa","family":"Head-Gordon","sequence":"additional","affiliation":[{"name":"University of California , , , ,","place":["Berkeley, California, United States, 94720"]},{"name":"University of California , , , ,","place":["Berkeley, California, United States, 94720"]},{"name":"University of California , , , ,","place":["Berkeley, California, United States, 94720"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"316","published-online":{"date-parts":[[2026,8,18]]},"reference":[{"issue":"8","key":"2026091409073419200_cit1","doi-asserted-by":"publisher","first-page":"660","DOI":"10.1038\/nrd1467","article-title":"Fragment-based lead\ndiscovery","volume":"3","author":"Rees","year":"2004","journal-title":"Nat. Rev. Drug Discovery"},{"issue":"14","key":"2026091409073419200_cit2","doi-asserted-by":"publisher","first-page":"3463","DOI":"10.1021\/jm040031v","article-title":"Fragment-Based\nDrug Discovery","volume":"47","author":"Erlanson","year":"2004","journal-title":"J. Med. Chem."},{"issue":"13","key":"2026091409073419200_cit3","doi-asserted-by":"publisher","first-page":"3661","DOI":"10.1021\/jm8000373","article-title":"Recent Developments\nin Fragment-Based Drug Discovery","volume":"51","author":"Congreve","year":"2008","journal-title":"J. Med. Chem."},{"issue":"9","key":"2026091409073419200_cit4","doi-asserted-by":"publisher","first-page":"605","DOI":"10.1038\/nrd.2016.109","article-title":"Twenty years\non: the impact of fragments on drug discovery","volume":"15","author":"Erlanson","year":"2016","journal-title":"Nat. Rev. Drug Discovery"},{"issue":"3","key":"2026091409073419200_cit5","doi-asserted-by":"publisher","first-page":"59","DOI":"10.1208\/s12248-018-0216-7","article-title":"Computational\nFragment-Based Drug Design: Current Trends,\nStrategies, and Applications","volume":"20","author":"Bian","year":"2018","journal-title":"Aaps J."},{"issue":"23","key":"2026091409073419200_cit6","doi-asserted-by":"publisher","first-page":"4309","DOI":"10.3390\/molecules24234309","article-title":"Concepts and Core Principles\nof Fragment-Based Drug Design","volume":"24","author":"Kirsch","year":"2019","journal-title":"Molecules"},{"issue":"20","key":"2026091409073419200_cit7","doi-asserted-by":"publisher","first-page":"11420","DOI":"10.1021\/acs.jmedchem.0c00242","article-title":"Fragment Linking\nStrategies for Structure-Based\nDrug Design","volume":"63","author":"Bancet","year":"2020","journal-title":"J. Med. Chem."},{"issue":"9","key":"2026091409073419200_cit8","doi-asserted-by":"publisher","first-page":"987","DOI":"10.1080\/17460441.2023.2234285","article-title":"Linkers in\nfragment-based drug design: an overview of the literature","volume":"18","author":"Grenier","year":"2023","journal-title":"Expert Opin. Drug Discovery"},{"issue":"4","key":"2026091409073419200_cit9","doi-asserted-by":"publisher","first-page":"1983","DOI":"10.1021\/acs.jcim.9b01120","article-title":"Deep Generative\nModels for 3D Linker Design","volume":"60","author":"Imrie","year":"2020","journal-title":"J. Chem. Inf. Model."},{"issue":"2","key":"2026091409073419200_cit10","doi-asserted-by":"publisher","first-page":"392","DOI":"10.1039\/D2DD00115B","article-title":"Link-INVENT: generative linker design with\nreinforcement learning","volume":"2","author":"Guo","year":"2023","journal-title":"Digital Discovery"},{"issue":"3","key":"2026091409073419200_cit11","doi-asserted-by":"publisher","first-page":"448","DOI":"10.1038\/s42256-025-00997-w","article-title":"Deep lead\noptimization enveloped in protein pocket\nand its application in designing potent and selective ligands targeting\nLTK protein","volume":"7","author":"Chen","year":"2025","journal-title":"Nat. Mach. Intell."},{"issue":"4","key":"2026091409073419200_cit12","doi-asserted-by":"publisher","first-page":"417","DOI":"10.1038\/s42256-024-00815-9","article-title":"Equivariant 3D-conditional\ndiffusion model for molecular linker design","volume":"6","author":"Igashov","year":"2024","journal-title":"Nat. Mach. Intell."},{"issue":"2","key":"2026091409073419200_cit13","doi-asserted-by":"publisher","first-page":"90","DOI":"10.1038\/nchem.1243","article-title":"Quantifying\nthe chemical beauty of drugs","volume":"4","author":"Bickerton","year":"2012","journal-title":"Nat. Chem."},{"issue":"1","key":"2026091409073419200_cit14","doi-asserted-by":"publisher","first-page":"8","DOI":"10.1186\/1758-2946-1-8","article-title":"Estimation\nof synthetic accessibility score of drug-like\nmolecules based on molecular complexity and fragment contributions","volume":"1","author":"Ertl","year":"2009","journal-title":"J. Cheminf."},{"key":"2026091409073419200_cit15","unstructured":"Walters, P.\n          \n          Generative Molecular\nDesign Isn\u2019t As Easy As People Make It Look; 2024. https:\/\/practicalcheminformatics.blogspot.com\/2024\/05\/generative-molecular-design-isnt-as.html."},{"key":"2026091409073419200_cit16","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2108.07258","article-title":"On\nthe Opportunities and\nRisks of Foundation Models","author":"Bommasani","year":"2022","journal-title":"arXiv"},{"issue":"7992","key":"2026091409073419200_cit17","doi-asserted-by":"publisher","first-page":"570","DOI":"10.1038\/s41586-023-06792-0","article-title":"Autonomous chemical\nresearch with large language models","volume":"624","author":"Boiko","year":"2023","journal-title":"Nature"},{"issue":"5","key":"2026091409073419200_cit18","doi-asserted-by":"publisher","first-page":"525","DOI":"10.1038\/s42256-024-00832-8","article-title":"Augmenting\nlarge language models with chemistry tools","volume":"6","author":"Bran","year":"2024","journal-title":"Nat. Mach. Intell."},{"key":"2026091409073419200_cit19","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2402.09391","article-title":"LlaSMol: Advancing Large Language Models for Chemistry\nwith a Large-Scale, Comprehensive, High-Quality Instruction Tuning\nDataset","author":"Yu","year":"2024","journal-title":"arXiv"},{"key":"2026091409073419200_cit20","doi-asserted-by":"publisher","first-page":"2514","DOI":"10.1039\/d4sc03921a","article-title":"A review of large language models\nand autonomous agents in chemistry","volume":"16","author":"Ramos","year":"2025","journal-title":"Chem. Sci."},{"key":"2026091409073419200_cit21","doi-asserted-by":"publisher","DOI":"10.26434\/chemrxiv-2025-hm3zb","article-title":"Exploring Transition Metal Complexes\nwith Large Language Models","author":"Liu","year":"2025","journal-title":"ChemRxiv"},{"key":"2026091409073419200_cit22","doi-asserted-by":"publisher","DOI":"10.1038\/s43588-026-00986-y","article-title":"SmileyLlama: modifying large\nlanguage\nmodels for directed chemical space exploration","author":"Cavanagh","year":"2026","journal-title":"Nat. Comput. Sci."},{"issue":"11","key":"2026091409073419200_cit23","doi-asserted-by":"publisher","first-page":"2108","DOI":"10.1021\/acscentsci.5c01285","article-title":"SynLlama: Generating Synthesizable Molecules and Their\nAnalogs with Large Language Models [Journal Article]","volume":"11","author":"Sun","year":"2025","journal-title":"ACS Cent. Sci."},{"issue":"1","key":"2026091409073419200_cit24","doi-asserted-by":"publisher","first-page":"1","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. Cheminform."},{"issue":"7","key":"2026091409073419200_cit25","doi-asserted-by":"publisher","DOI":"10.1126\/sciadv.aap7885","article-title":"Deep reinforcement\nlearning for de\nnovo drug design","volume":"4","author":"Popova","year":"2018","journal-title":"Sci. Adv."},{"key":"2026091409073419200_cit26","doi-asserted-by":"publisher","first-page":"D1100","DOI":"10.1093\/nar\/gkr777","article-title":"ChEMBL: a large-scale\nbioactivity database for drug\ndiscovery","volume":"40","author":"Gaulton","year":"2012","journal-title":"Nucleic Acids Res."},{"key":"2026091409073419200_cit27","unstructured":"rdkit\/rdkit.\n          \n          RDKit; 2026. https:\/\/github.com\/rdkit\/rdkit."},{"key":"2026091409073419200_cit28","doi-asserted-by":"publisher","first-page":"9082","DOI":"10.1021\/acs.jcim.4c00634","article-title":"Mining\nfor Potent Inhibitors through Artificial Intelligence\nand Physics: A Unified Methodology for Ligand Based and Structure\nBased Drug Design","volume":"64","author":"Li","year":"2024","journal-title":"J. Chem. Inf. Model."},{"key":"2026091409073419200_cit29","unstructured":"Walters, P.\n          \n          Mining Ring Systems in\nMolecules for Fun and Profit. https:\/\/practicalcheminformatics.blogspot.com\/2022\/12\/identifying-ring-systems-in-molecules.html."},{"issue":"21","key":"2026091409073419200_cit30","doi-asserted-by":"publisher","first-page":"22756","DOI":"10.1021\/acs.jmedchem.5c01706","article-title":"Structure-Based\nGeneration of 3D Small-Molecule Drugs: Are We There Yet?","volume":"68","author":"Yang","year":"2025","journal-title":"J. Med. Chem."},{"key":"2026091409073419200_cit31","unstructured":"Weininger, D.\n          \n          Daylight Theory: sMARTS\n- A Language for Describing Molecular Patterns. https:\/\/www.daylight.com\/dayhtml\/doc\/theory\/theory.smarts.html."},{"issue":"7","key":"2026091409073419200_cit32","doi-asserted-by":"publisher","first-page":"2719","DOI":"10.1021\/jm901137j","article-title":"New Substructure\nFilters for Removal of 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."},{"issue":"3","key":"2026091409073419200_cit33","doi-asserted-by":"publisher","first-page":"435","DOI":"10.1002\/cmdc.200700139","article-title":"Lessons\nLearnt from\nAssembling Screening Libraries for Drug Discovery for Neglected Diseases","volume":"3","author":"Brenk","year":"2008","journal-title":"ChemMedChem"},{"issue":"7","key":"2026091409073419200_cit34","doi-asserted-by":"publisher","first-page":"1757","DOI":"10.1021\/ci3001277","article-title":"ZINC: A\nFree Tool to Discover Chemistry for Biology","volume":"52","author":"Irwin","year":"2012","journal-title":"J. Chem. Inf. Model."},{"key":"2026091409073419200_cit35","unstructured":"openadmet (Open ADMET Consortium).\n          \n          OpenADMET; 2026. https:\/\/huggingface.co\/openadmet."},{"issue":"18","key":"2026091409073419200_cit36","doi-asserted-by":"publisher","first-page":"184110","DOI":"10.1063\/5.0297006","article-title":"A foundation model for atomistic materials\nchemistry","volume":"163","author":"Batatia","year":"2025","journal-title":"J. Chem. Phys."},{"issue":"1","key":"2026091409073419200_cit37","doi-asserted-by":"publisher","first-page":"11","DOI":"10.1038\/s41597-022-01882-6","article-title":"SPICE, A Dataset of\nDrug-like Molecules and Peptides for Training Machine Learning Potentials","volume":"10","author":"Eastman","year":"2023","journal-title":"Sci. Data"},{"key":"2026091409073419200_cit38","doi-asserted-by":"publisher","first-page":"6613","DOI":"10.1021\/acs.jcim.5c00586","article-title":"Strain\nProblems Got You in a Twist?\nTry StrainRelief: A Quantum-Accurate Tool for Ligand Strain Calculations","volume":"65","author":"Wallace","year":"2025","journal-title":"J. Chem. Inf. Model."},{"issue":"5","key":"2026091409073419200_cit39","doi-asserted-by":"publisher","first-page":"1209","DOI":"10.1039\/d4dd00357h","article-title":"A workflow\nto create a high-quality protein\u2013ligand binding dataset for\ntraining, validation, and prediction tasks","volume":"4","author":"Wang","year":"2025","journal-title":"Digital Discovery"},{"issue":"11","key":"2026091409073419200_cit40","doi-asserted-by":"publisher","first-page":"3336","DOI":"10.1021\/acs.jctc.2c01145","article-title":"Uni-Dock:\nGPU-Accelerated Docking Enables Ultralarge Virtual Screening","volume":"19","author":"Yu","year":"2023","journal-title":"J. Chem. Theory Comput."},{"issue":"10","key":"2026091409073419200_cit41","doi-asserted-by":"publisher","first-page":"1615","DOI":"10.1002\/1873-3468.13746","article-title":"Crystal\nstructure of the mineralocorticoid receptor\nligand\u2013binding domain in complex with a potent and selective\nnonsteroidal blocker, esaxerenone (CS\u20133150)","volume":"594","author":"Takahashi","year":"2020","journal-title":"FEBS Lett."},{"issue":"3","key":"2026091409073419200_cit42","doi-asserted-by":"publisher","first-page":"181","DOI":"10.1038\/s41573-021-00371-6","article-title":"PROTAC\ntargeted protein degraders:\nthe past is prologue","volume":"21","author":"Be\u0301ke\u0301s","year":"2022","journal-title":"Nat. Rev. Drug Discovery"},{"issue":"D1","key":"2026091409073419200_cit43","doi-asserted-by":"publisher","first-page":"D1510","DOI":"10.1093\/nar\/gkae768","article-title":"PROTAC-DB\n3.0: an updated database of PROTACs with extended pharmacokinetic\nparameters","volume":"53","author":"Ge","year":"2025","journal-title":"Nucleic Acids Res."},{"issue":"4","key":"2026091409073419200_cit44","doi-asserted-by":"publisher","first-page":"1727","DOI":"10.1002\/anie.201914396","article-title":"Structure-Based\nDesign of a Macrocyclic PROTAC","volume":"59","author":"Testa","year":"2020","journal-title":"Angew. Chem.,\nInt. Ed."},{"key":"2026091409073419200_cit45","unstructured":"Home | openbind.uk. https:\/\/openbind.uk\/."},{"issue":"6","key":"2026091409073419200_cit46","doi-asserted-by":"publisher","first-page":"3129","DOI":"10.1021\/acs.jcim.5c02106","article-title":"A Computational\nCommunity Blind Challenge on\nPan-Coronavirus Drug Discovery Data","volume":"66","author":"MacDermott-Opeskin","year":"2026","journal-title":"J. Chem.\nInf. Model."},{"issue":"21","key":"2026091409073419200_cit47","doi-asserted-by":"publisher","first-page":"12058","DOI":"10.1021\/acs.jcim.5c01976","article-title":"Polaris Challenge:\nData-Driven Priors to Improve Docking for Pose Prediction","volume":"65","author":"Sun","year":"2025","journal-title":"J. Chem. Inf. Model."},{"key":"2026091409073419200_cit48","article-title":"Accelerating Scientific\nDiscovery\nwith Autonomous Goal-evolving Agents","author":"Du","journal-title":"arXiv"},{"issue":"1","key":"2026091409073419200_cit49","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."}],"container-title":["Journal of Chemical Information\nand Modeling"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/pubs.acs.org\/doi\/pdf\/10.1021\/acs.jcim.6c01261","content-type":"application\/pdf","content-version":"vor","intended-application":"unspecified"},{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article-pdf\/66\/17\/11348\/66826920\/acs.jcim.6c01261.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article-pdf\/66\/17\/11348\/66826920\/acs.jcim.6c01261.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,14]],"date-time":"2026-09-14T13:21:20Z","timestamp":1789392080000},"score":1,"resource":{"primary":{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article\/66\/17\/11348\/5273939\/LinkLlama-Enabling-a-Large-Language-Model-for"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,18]]},"references-count":49,"journal-issue":{"issue":"17","published-online":{"date-parts":[[2026,8,18]]},"published-print":{"date-parts":[[2026,9,14]]}},"URL":"https:\/\/doi.org\/10.1021\/acs.jcim.6c01261","relation":{},"ISSN":["1549-9596","1549-960X"],"issn-type":[{"value":"1549-9596","type":"print"},{"value":"1549-960X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,18]]}}}