{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,18]],"date-time":"2026-09-18T06:03:29Z","timestamp":1789711409361,"version":"4.0.1"},"reference-count":100,"publisher":"American Association for the Advancement of Science (AAAS)","content-domain":{"domain":["www.science.org"],"crossmark-restriction":true},"short-container-title":["Science"],"abstract":"<jats:p>Drug development requires evidence integration across biological scales and modalities, but relevant tools are fragmented. We introduce the Virtual Biotech, an organization of artificial intelligence (AI) agents modeled on a drug-development company, with agentic divisions spanning target discovery, safety assessment, modality selection, and clinical development. We demonstrate its utility at three drug-development decision points. First, over 37,000 agents annotated outcomes from 55,984 trials and found that drugs targeting cell-type-specific genes were 48% more likely to reach market with 32% fewer adverse events. Second, it integrated multimodal evidence to propose a therapeutic strategy in lung cancer. Third, it analyzed a terminated ulcerative colitis trial and inferred potential mechanisms of failure. These results demonstrate that human-guided multi-agent systems can conduct transparent, multiscale analyses to inform therapeutic-development decisions.<\/jats:p>","DOI":"10.1126\/science.aeg6779","type":"journal-article","created":{"date-parts":[[2026,9,17]],"date-time":"2026-09-17T18:00:19Z","timestamp":1789668019000},"update-policy":"https:\/\/doi.org\/10.34133\/aaas_crossmark","source":"Crossref","is-referenced-by-count":1,"title":["The Virtual Biotech: A multi-agent AI framework for therapeutic discovery and development"],"prefix":"10.1126","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-6679-1464","authenticated-orcid":true,"given":"Harrison G.","family":"Zhang","sequence":"first","affiliation":[{"name":"Department of Biomedical Data Science, Stanford University, Stanford, CA, USA."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5388-9451","authenticated-orcid":true,"given":"Peter","family":"Eckmann","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Stanford University, Stanford, CA, USA."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4524-7408","authenticated-orcid":true,"given":"Jiacheng","family":"Miao","sequence":"additional","affiliation":[{"name":"Department of Biomedical Data Science, Stanford University, Stanford, CA, USA."},{"name":"Department of Genetics, Stanford University, Stanford, CA, USA."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5842-3944","authenticated-orcid":true,"given":"Andrew B.","family":"Mahon","sequence":"additional","affiliation":[{"name":"PHD Biosciences, La Jolla, CA, USA."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8880-4764","authenticated-orcid":true,"given":"James","family":"Zou","sequence":"additional","affiliation":[{"name":"Department of Biomedical Data Science, Stanford University, Stanford, CA, USA."},{"name":"Department of Computer Science, Stanford University, Stanford, CA, USA."}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"221","published-online":{"date-parts":[[2026,9,17]]},"reference":[{"key":"e_1_3_2_2_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.jhealeco.2016.01.012"},{"key":"e_1_3_2_3_2","doi-asserted-by":"publisher","DOI":"10.1001\/jama.2020.1166"},{"key":"e_1_3_2_4_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.apsb.2022.02.002"},{"key":"e_1_3_2_5_2","doi-asserted-by":"publisher","DOI":"10.1038\/nbt.2786"},{"key":"e_1_3_2_6_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-025-64552-2"},{"key":"e_1_3_2_7_2","doi-asserted-by":"publisher","DOI":"10.1093\/biostatistics\/kxx069"},{"key":"e_1_3_2_8_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.conctc.2018.08.001"},{"key":"e_1_3_2_9_2","doi-asserted-by":"publisher","DOI":"10.1146\/annurev-biodatasci-102523-103838"},{"key":"e_1_3_2_10_2","doi-asserted-by":"publisher","DOI":"10.1038\/nrd1608"},{"key":"e_1_3_2_11_2","doi-asserted-by":"publisher","DOI":"10.1186\/s13023-021-01806-4"},{"key":"e_1_3_2_12_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.tips.2015.12.001"},{"key":"e_1_3_2_13_2","doi-asserted-by":"publisher","DOI":"10.2196\/51297"},{"key":"e_1_3_2_14_2","unstructured":"S. Han Q. Zhang Y. Yao W. Jin Z. Xu LLM multi-agent systems: Challenges and open problems. arXiv:2402.03578 [cs.MA] (2024)."},{"key":"e_1_3_2_15_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-023-06792-0"},{"key":"e_1_3_2_16_2","doi-asserted-by":"publisher","DOI":"10.1038\/s42256-024-00832-8"},{"key":"e_1_3_2_17_2","unstructured":"C. Lu. The AI scientist: Towards fully automated open-ended scientific discovery. arXiv:2408.06292 [cs.AI] (2024)."},{"key":"e_1_3_2_18_2","doi-asserted-by":"crossref","unstructured":"J. Miao J. R. Davis Y. Zhang J. K. Pritchard J. Zou Paper2Agent: Reimagining research papers as interactive and reliable AI agents. arXiv:2509.06917 [cs.AI] (2025).","DOI":"10.1038\/s41586-026-11044-y"},{"key":"e_1_3_2_19_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41551-025-01463-z"},{"key":"e_1_3_2_20_2","doi-asserted-by":"crossref","unstructured":"H. Wang et al. SpatialAgent: An autonomous AI agent for spatial biology. bioRxiv 2025.04.03.646459 [Preprint] (2025); https:\/\/doi.org\/10.1101\/2025.04.03.646459.","DOI":"10.1101\/2025.04.03.646459"},{"key":"e_1_3_2_21_2","doi-asserted-by":"crossref","unstructured":"S. Alber et al. CellVoyager: AI compbio agent generates new insights by autonomously analyzing biological data. bioRxiv 2025.06.03.657517 [Preprint] (2025); https:\/\/doi.org\/10.1101\/2025.06.03.657517.","DOI":"10.1101\/2025.06.03.657517"},{"key":"e_1_3_2_22_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-025-10097-9"},{"key":"e_1_3_2_23_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-025-09442-9"},{"key":"e_1_3_2_24_2","doi-asserted-by":"crossref","unstructured":"K. Huang et al. Biomni: A general-purpose biomedical AI agent. bioRxiv 2025.05.30.656746 [Preprint] (2025); https:\/\/doi.org\/10.1101\/2025.05.30.656746.","DOI":"10.1101\/2025.05.30.656746"},{"key":"e_1_3_2_25_2","unstructured":"L. Mitchener. Kosmos: An AI scientist for autonomous discovery. arXiv:2511.02824 [cs.AI] (2025)."},{"key":"e_1_3_2_26_2","unstructured":"W. Xu et al. PantheonOS: An evolvable multi-agent framework for automatic genomics discovery. bioRxiv 2026.02.26.707870 [Preprint] (2026); 2026.02.26.707870."},{"key":"e_1_3_2_27_2","doi-asserted-by":"crossref","unstructured":"R. Jin et al. STELLA: Towards a biomedical world model with self-evolving multimodal agents. bioRxiv \u00a0 2025.07.01.662467 [Preprint] (2025); https:\/\/doi.org\/10.1101\/2025.07.01.662467.","DOI":"10.1101\/2025.07.01.662467"},{"key":"e_1_3_2_28_2","unstructured":"B. Gao. PharmAgents: Building a virtual pharma with large language model agents. arXiv:2503.22164 [q-bio.BM] (2025)."},{"key":"e_1_3_2_29_2","doi-asserted-by":"crossref","unstructured":"P. Sui et al. Medea: An omics AI agent for therapeutic discovery. bioRxiv 2026.01.16.696667 [Preprint] (2026); https:\/\/doi.org\/10.64898\/2026.01.16.696667.","DOI":"10.64898\/2026.01.16.696667"},{"key":"e_1_3_2_30_2","unstructured":"Anthropic Introducing the Model Context Protocol (2024). https:\/\/www.anthropic.com\/news\/model-context-protocol."},{"key":"e_1_3_2_31_2","unstructured":"What is the Model Context Protocol (MCP)? (2026); https:\/\/modelcontextprotocol.io\/docs\/getting-started\/intro."},{"key":"e_1_3_2_32_2","doi-asserted-by":"crossref","unstructured":"X. Hou Y. Zhao S. Wang H. Wang Model context protocol (MCP): Landscape security threats and future research directions. arXiv:2503.23278 [cs.CR] (2025).","DOI":"10.1145\/3796519"},{"key":"e_1_3_2_33_2","unstructured":"Open Targets 25.09 Platform release now live! (2025)."},{"key":"e_1_3_2_34_2","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkae1128"},{"key":"e_1_3_2_35_2","doi-asserted-by":"crossref","unstructured":"J. Zhang et al. Tahoe-100M: A giga-scale single-cell perturbation atlas for context-dependent gene function and cellular modeling. bioRxiv 2025.02.20.639398 [Preprint] (2025); https:\/\/doi.org\/10.1101\/2025.02.20.639398.","DOI":"10.1101\/2025.02.20.639398"},{"key":"e_1_3_2_36_2","doi-asserted-by":"publisher","DOI":"10.1136\/jamia.2000.0070313"},{"key":"e_1_3_2_37_2","doi-asserted-by":"crossref","unstructured":"S. Wu. CollabLLM: From passive responders to active collaborators. arXiv:2502.00640 [cs.AI] (2025).","DOI":"10.24963\/ijcai.2026\/928"},{"key":"e_1_3_2_38_2","unstructured":"L. Kuhn Y. Gal S. Farquhar CLAM: Selective clarification for ambiguous questions with generative language models. arXiv:2212.07769 [cs.CL] (2022)."},{"key":"e_1_3_2_39_2","doi-asserted-by":"publisher","DOI":"10.1038\/d41573-019-00074-z"},{"key":"e_1_3_2_40_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-024-07316-0"},{"key":"e_1_3_2_41_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41588-024-01854-z"},{"key":"e_1_3_2_42_2","doi-asserted-by":"publisher","DOI":"10.1038\/d41573-022-00120-3"},{"key":"e_1_3_2_43_2","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkw1055"},{"key":"e_1_3_2_44_2","doi-asserted-by":"publisher","DOI":"10.1136\/bmj.k1452"},{"key":"e_1_3_2_45_2","doi-asserted-by":"publisher","DOI":"10.7326\/L14-5022"},{"key":"e_1_3_2_46_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.patter.2022.100445"},{"key":"e_1_3_2_47_2","doi-asserted-by":"crossref","unstructured":"S. R. Quake The Tabula Sapiens Consortium Tabula sapiens reveals transcription factor expression senescence effects and sex-specific features in cell types from 28 human organs and tissues. bioRxiv 2024.12.03.626516 [Preprint] (2025); https:\/\/doi.org\/10.1101\/2024.12.03.626516.","DOI":"10.1101\/2024.12.03.626516"},{"key":"e_1_3_2_48_2","first-page":"205","article-title":"A benchmark of gene expression tissue-specificity metrics","volume":"18","author":"Kryuchkova-Mostacci N.","year":"2017","unstructured":"N. Kryuchkova-Mostacci, M. Robinson-Rechavi, A benchmark of gene expression tissue-specificity metrics. Brief. Bioinform. 18, 205\u2013214 (2017). 26891983","journal-title":"Brief. Bioinform."},{"key":"e_1_3_2_49_2","doi-asserted-by":"publisher","DOI":"10.1214\/aos\/1176346577"},{"key":"e_1_3_2_50_2","doi-asserted-by":"publisher","DOI":"10.3389\/fpsyg.2013.00700"},{"key":"e_1_3_2_51_2","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/bts713"},{"key":"e_1_3_2_52_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.yjbinx.2018.100001"},{"key":"e_1_3_2_53_2","doi-asserted-by":"publisher","DOI":"10.1186\/s13059-016-1077-y"},{"key":"e_1_3_2_54_2","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0159641"},{"key":"e_1_3_2_55_2","doi-asserted-by":"publisher","DOI":"10.1093\/poq\/nfw005"},{"key":"e_1_3_2_56_2","doi-asserted-by":"publisher","DOI":"10.3758\/s13428-012-0225-x"},{"key":"e_1_3_2_57_2","unstructured":"E. Dann et al. Estimating the impact of single-cell RNA sequencing of human tissues on drug target validation. medRxiv 2024.04.04.24305313 [Preprint] (2024); https:\/\/doi.org\/10.1101\/2024.04.04.24305313.10.1101\/2024.04.04.24305313"},{"key":"e_1_3_2_58_2","doi-asserted-by":"publisher","DOI":"10.3389\/fimmu.2021.701006"},{"key":"e_1_3_2_59_2","doi-asserted-by":"publisher","DOI":"10.1158\/1078-0432.CCR-20-2584"},{"key":"e_1_3_2_60_2","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkae1142"},{"key":"e_1_3_2_61_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41556-024-01469-w"},{"key":"e_1_3_2_62_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-024-48700-8"},{"key":"e_1_3_2_63_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41587-021-01139-4"},{"key":"e_1_3_2_64_2","doi-asserted-by":"publisher","DOI":"10.1126\/scisignal.2004088"},{"key":"e_1_3_2_65_2","doi-asserted-by":"publisher","DOI":"10.1038\/ng.2764"},{"key":"e_1_3_2_66_2","doi-asserted-by":"publisher","DOI":"10.1038\/nbt1210-1248"},{"key":"e_1_3_2_67_2","unstructured":"Merck Ifinatamab deruxtecan granted breakthrough therapy designation by U.S. FDA for patients with pretreated extensive-stage small cell lung cancer (2025); https:\/\/www.merck.com\/news\/ifinatamab-deruxtecan-granted-breakthrough-therapy-designation-by-u-s-fda-for-patients-with-pretreated-extensive-stage-small-cell-lung-cancer\/."},{"key":"e_1_3_2_68_2","unstructured":"Merck Ifinatamab deruxtecan demonstrated clinically meaningful response rates in patients with extensive-stage small cell lung cancer in IDeate-Lung01 phase 2 trial (2025); https:\/\/www.merck.com\/news\/ifinatamab-deruxtecan-demonstrated-clinically-meaningful-response-rates-in-patients-with-extensive-stage-small-cell-lung-cancer-in-ideate-lung01-phase-2-trial\/."},{"key":"e_1_3_2_69_2","doi-asserted-by":"publisher","DOI":"10.1200\/JCO-25-02142"},{"key":"e_1_3_2_70_2","unstructured":"ClinicalTrials.gov A study to evaluate the efficacy safety and pharmacokinetics of vixarelimab in participants with moderate to severe ulcerative colitis (UC) (2026); https:\/\/clinicaltrials.gov\/study\/NCT06137183."},{"key":"e_1_3_2_71_2","doi-asserted-by":"publisher","DOI":"10.3389\/fimmu.2023.1239732"},{"key":"e_1_3_2_72_2","doi-asserted-by":"publisher","DOI":"10.1038\/nm.4307"},{"key":"e_1_3_2_73_2","doi-asserted-by":"publisher","DOI":"10.1093\/ibd\/izab032"},{"key":"e_1_3_2_74_2","doi-asserted-by":"publisher","DOI":"10.1007\/s00535-021-01819-7"},{"key":"e_1_3_2_75_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41590-024-01994-8"},{"key":"e_1_3_2_76_2","doi-asserted-by":"publisher","DOI":"10.1093\/bioadv\/vbac016"},{"key":"e_1_3_2_77_2","doi-asserted-by":"publisher","DOI":"10.1093\/ecco-jcc\/jjae200"},{"key":"e_1_3_2_78_2","doi-asserted-by":"publisher","DOI":"10.1136\/gut.2009.178665"},{"key":"e_1_3_2_79_2","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0007984"},{"key":"e_1_3_2_80_2","doi-asserted-by":"publisher","DOI":"10.1038\/ajg.2011.83"},{"key":"e_1_3_2_81_2","doi-asserted-by":"publisher","DOI":"10.1136\/gutjnl-2016-312293"},{"key":"e_1_3_2_82_2","doi-asserted-by":"publisher","DOI":"10.3390\/ijms23031811"},{"key":"e_1_3_2_83_2","doi-asserted-by":"publisher","DOI":"10.7554\/eLife.104815.3"},{"key":"e_1_3_2_84_2","doi-asserted-by":"publisher","DOI":"10.1186\/s12885-026-15647-6"},{"key":"e_1_3_2_85_2","unstructured":"Anthropic Agent SDK overview (2026); https:\/\/platform.claude.com\/docs\/en\/agent-sdk\/overview."},{"key":"e_1_3_2_86_2","doi-asserted-by":"publisher","DOI":"10.1093\/jamiaopen\/ooae146"},{"key":"e_1_3_2_87_2","unstructured":"ClinicalTrials.gov API migration guide (info API endpoints) (2024); https:\/\/clinicaltrials.gov\/data-api\/about-api\/api-migration#info-endpoints."},{"key":"e_1_3_2_88_2","unstructured":"Prefect Welcome to FastMCP (2026); https:\/\/gofastmcp.com\/getting-started\/welcome."},{"key":"e_1_3_2_89_2","unstructured":"National Center for Biotechnology Information About \u2013 PubMed (2025); https:\/\/pubmed.ncbi.nlm.nih.gov\/about\/."},{"key":"e_1_3_2_90_2","doi-asserted-by":"publisher","DOI":"10.18637\/jss.v034.i02"},{"key":"e_1_3_2_91_2","doi-asserted-by":"publisher","DOI":"10.1111\/j.2517-6161.1995.tb02031.x"},{"key":"e_1_3_2_92_2","doi-asserted-by":"publisher","DOI":"10.1186\/s13059-017-1382-0"},{"key":"e_1_3_2_93_2","doi-asserted-by":"publisher","DOI":"10.1186\/s13326-016-0088-7"},{"key":"e_1_3_2_94_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41592-019-0619-0"},{"key":"e_1_3_2_95_2","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btad547"},{"key":"e_1_3_2_96_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41592-019-0686-2"},{"key":"e_1_3_2_97_2","doi-asserted-by":"publisher","DOI":"10.18637\/jss.v082.i13"},{"key":"e_1_3_2_98_2","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkad841"},{"key":"e_1_3_2_99_2","volume-title":"Introduction to Bivariate and Multivariate Analysis","author":"Lindeman R. H.","year":"1980","unstructured":"R. H. Lindeman, P. F. Merenda, R. Z. Gold, Introduction to Bivariate and Multivariate Analysis, vol. 4 (Scott, 1980)."},{"key":"e_1_3_2_100_2","unstructured":"K. Barton MuMIn: Multi-Model Inference (R package version 1.18 2015)."},{"key":"e_1_3_2_101_2","unstructured":"H. G. Zhang P. Eckmann J. Miao A. B. Mahon J. Zou The Virtual Biotech: Case Study Documentation Zenodo (2026); https:\/\/doi.org\/10.5281\/zenodo.22259123."}],"container-title":["Science"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.science.org\/doi\/pdf\/10.1126\/science.aeg6779","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,17]],"date-time":"2026-09-17T18:00:29Z","timestamp":1789668029000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.science.org\/doi\/10.1126\/science.aeg6779"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,17]]},"references-count":100,"alternative-id":["10.1126\/science.aeg6779"],"URL":"https:\/\/doi.org\/10.1126\/science.aeg6779","relation":{},"ISSN":["0036-8075","1095-9203"],"issn-type":[{"value":"0036-8075","type":"print"},{"value":"1095-9203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,17]]},"assertion":[{"value":"2026-02-23","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-05-25","order":1,"name":"revised","label":"Revised","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-09-02","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-09-17","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}],"article-number":"eaeg6779"}}