{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,16]],"date-time":"2026-09-16T18:03:42Z","timestamp":1789581822161,"version":"build-2803163510"},"reference-count":27,"publisher":"Public Library of Science (PLoS)","issue":"9","license":[{"start":{"date-parts":[[2026,9,16]],"date-time":"2026-09-16T00:00:00Z","timestamp":1789516800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["2025457"],"award-info":[{"award-number":["2025457"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100009612","name":"Oregon State University","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100009612","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["www.ploscompbiol.org"],"crossmark-restriction":false},"short-container-title":["PLoS Comput Biol"],"abstract":"<jats:p>\n                    Responsible use of large language models (LLMs) in life science research demands a unified approach to risk, yet existing guidance treats prompting and verification as separate topics rather than as integrated components of a single risk management framework. This paper addresses that gap for life science researchers. We explain how LLM architecture produces both remarkable capabilities and characteristic failures including hallucination and sycophancy, with particular attention to vulnerabilities most relevant to life science workflows. We argue that effective prompt design is itself a primary risk management strategy, reducing the probability and severity of characteristic failures before verification is required, and introduce a structured framework for matching downstream verification effort to risk across dimensions of output verifiability, researcher expertise, and consequence of error. We apply both layers to common research tasks, including literature synthesis, code generation, writing assistance, statistical reasoning, and administrative work, with documentation practices and ethical obligations addressed throughout. Six supplementary guides extend each component into detailed workflows, worked examples, and reference checklists, and a companion open-source repository (\n                    <jats:ext-link xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" ext-link-type=\"uri\" xlink:href=\"https:\/\/github.com\/SharptonLab\/PromptLab\" xlink:type=\"simple\">https:\/\/github.com\/SharptonLab\/PromptLab<\/jats:ext-link>\n                    ) provides tested prompts and verification tools for immediate use. Together, these resources are designed to help researchers use these tools carefully and document how they did so.\n                  <\/jats:p>","DOI":"10.1371\/journal.pcbi.1014776","type":"journal-article","created":{"date-parts":[[2026,9,16]],"date-time":"2026-09-16T17:33:58Z","timestamp":1789580038000},"page":"e1014776","update-policy":"https:\/\/doi.org\/10.1371\/journal.pcbi.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["A practical risk framework for large language model use in life science research"],"prefix":"10.1371","volume":"22","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-5996-7764","authenticated-orcid":true,"given":"Thomas","family":"Sharpton","sequence":"first","affiliation":[{"name":"Linus Pauling Institute, Oregon State University","place":["Corvallis, Oregon, United States of America"]},{"name":"Oregon State University","place":["Corvallis, Oregon, United States of America"],"department":["Department of Microbiology"]},{"name":"Oregon State University","place":["Corvallis, Oregon, United States of America"],"department":["Department of Statistics"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Edward W.","family":"Davis II","sequence":"additional","affiliation":[{"name":"Oregon State University","place":["Corvallis, Oregon, United States of America"],"department":["Center for Quantitative Life Sciences"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alexandra","family":"Alexiev","sequence":"additional","affiliation":[{"name":"Linus Pauling Institute, Oregon State University","place":["Corvallis, Oregon, United States of America"]},{"name":"Oregon State University","place":["Corvallis, Oregon, United States of America"],"department":["Department of Microbiology"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,9,16]]},"reference":[{"key":"pcbi.1014776.ref001","author":"Z Liao","year":"2024"},{"key":"pcbi.1014776.ref002","doi-asserted-by":"crossref","first-page":"672","DOI":"10.1038\/d41586-023-02980-0","article-title":"AI and science: What 1,600 researchers think","volume":"621","author":"R Van Noorden","year":"2023","journal-title":"Nature"},{"issue":"12","key":"pcbi.1014776.ref003","doi-asserted-by":"crossref","first-page":"2629","DOI":"10.1007\/s10439-023-03272-4","article-title":"Prompt Engineering with ChatGPT: a guide for academic writers","volume":"51","author":"L Giray","year":"2023","journal-title":"Ann Biomed Eng"},{"key":"pcbi.1014776.ref004","author":"S Schulhoff","year":"2024"},{"key":"pcbi.1014776.ref005","doi-asserted-by":"crossref","first-page":"1012","DOI":"10.1111\/2041-210x.70267","article-title":"Prompting LLMs for quality ecological statistics","volume":"17","author":"CJ Brown","year":"2026","journal-title":"Methods Ecol Evol"},{"key":"pcbi.1014776.ref006","unstructured":"Hendrycks D, Burns C, Basart S, Zou A, Mazeika M, Song D, et al. Measuring massive multitask language understanding. In: ICLR 2021, 2021. https:\/\/doi.org\/arXiv:2009.03300"},{"key":"pcbi.1014776.ref007","unstructured":"Rein D, Hou BL, Stickland AC, et al. GPQA: A graduate-level Google-proof Q&A benchmark. In: COLM 2024, 2024. https:\/\/doi.org\/arXiv:2311.12022"},{"issue":"7945","key":"pcbi.1014776.ref008","doi-asserted-by":"crossref","first-page":"612","DOI":"10.1038\/d41586-023-00191-1","article-title":"Tools such as ChatGPT threaten transparent science; here are our ground rules for their use","volume":"613","year":"2023","journal-title":"Nature"},{"issue":"4","key":"pcbi.1014776.ref009","doi-asserted-by":"crossref","first-page":"449","DOI":"10.1177\/17470161231180449","article-title":"The ethics of disclosing the use of artificial intelligence tools in writing scholarly manuscripts","volume":"19","author":"M Hosseini","year":"2023","journal-title":"Res Ethics"},{"issue":"1","key":"pcbi.1014776.ref010","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pcbi.1011767","article-title":"Ten simple rules for using LLMs in science","volume":"20","author":"GR Smith","year":"2024","journal-title":"PLOS Comput Biol"},{"key":"pcbi.1014776.ref011","doi-asserted-by":"crossref","first-page":"157","DOI":"10.1162\/tacl_a_00638","article-title":"Lost in the middle: how language models use long contexts","volume":"12","author":"NF Liu","year":"2024","journal-title":"TACL"},{"key":"pcbi.1014776.ref012","unstructured":"Sharma M, Tong M, Korbak T. Towards understanding sycophancy in language models. In: ICLR 2024, 2024. https:\/\/doi.org\/arXiv:2310.13548"},{"key":"pcbi.1014776.ref013","doi-asserted-by":"crossref","unstructured":"Perez E, Ringer S, Lukosiute K, Nguyen K, Chen E, Heiner S, et al. Discovering language model behaviors with model-written evaluations. In: Findings of the Association for Computational Linguistics: ACL 2023, 2023. p. 13387\u2013434. https:\/\/doi.org\/10.18653\/v1\/2023.findings-acl.847","DOI":"10.18653\/v1\/2023.findings-acl.847"},{"issue":"12","key":"pcbi.1014776.ref014","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3571730","article-title":"Survey of hallucination in natural language generation","volume":"55","author":"Z Ji","year":"2023","journal-title":"ACM Comput Surv"},{"issue":"2","key":"pcbi.1014776.ref015","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3703155","article-title":"A survey on hallucination in LLMs: principles, taxonomy, challenges, and open questions","volume":"43","author":"L Huang","year":"2025","journal-title":"ACM Trans Inf Syst"},{"key":"pcbi.1014776.ref016","doi-asserted-by":"crossref","DOI":"10.52202\/068431-1800","article-title":"Chain-of-thought prompting elicits reasoning in LLMs","volume":"35","author":"J Wei","year":"2022","journal-title":"Adv Neural Information Process Syst"},{"key":"pcbi.1014776.ref017","doi-asserted-by":"crossref","unstructured":"Panickssery A, Bowman S, Feng S. LLM evaluators recognize and favor their own generations. In: Advances in Neural Information Processing Systems 37, 2024. 68772\u2013802. https:\/\/doi.org\/10.52202\/079017-2197","DOI":"10.52202\/079017-2197"},{"key":"pcbi.1014776.ref018","doi-asserted-by":"crossref","unstructured":"Cohen R, Hamri M, Geva M, Globerson A. LM vs LM: detecting factual errors via cross examination. In: Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, 2023. 12621\u201340. https:\/\/doi.org\/10.18653\/v1\/2023.emnlp-main.778","DOI":"10.18653\/v1\/2023.emnlp-main.778"},{"key":"pcbi.1014776.ref019","article-title":"Language models are few-shot learners","volume":"33","author":"TB Brown","year":"2020","journal-title":"Adv Neural Information Process Syst"},{"key":"pcbi.1014776.ref020","doi-asserted-by":"crossref","first-page":"14045","DOI":"10.1038\/s41598-023-41032-5","article-title":"Fabrication and errors in the bibliographic citations generated by ChatGPT","volume":"13","author":"WH Walters","year":"2023","journal-title":"Sci Rep"},{"issue":"1","key":"pcbi.1014776.ref021","first-page":"80","article-title":"ChatGPT hallucinates non-existent citations: evidence from economics","volume":"69","author":"J Buchanan","year":"2023","journal-title":"Am Econ"},{"key":"pcbi.1014776.ref022","article-title":"The impact of AI on developer productivity: evidence from GitHub Copilot","author":"S Peng","year":"2023","journal-title":"arXiv: 2302"},{"issue":"1","key":"pcbi.1014776.ref023","doi-asserted-by":"crossref","DOI":"10.1002\/smr.2723","article-title":"LLMs for science: usage for code generation and data analysis","volume":"37","author":"M Nejjar","year":"2024","journal-title":"J Software Evolu Process"},{"key":"pcbi.1014776.ref024","unstructured":"Nature Portfolio. Artificial intelligence (AI). Nature Portfolio Editorial Policies. 2024. Available from: https:\/\/www.nature.com\/nature-portfolio\/editorial-policies\/ai"},{"key":"pcbi.1014776.ref025","unstructured":"PLOS. Ethical publishing practice. PLOS ONE Editorial Policies. 2024. Available from: https:\/\/journals.plos.org\/plosone\/s\/ethical-publishing-practice"},{"key":"pcbi.1014776.ref026","unstructured":"Elsevier. The use of generative AI and AI-assisted technologies in the editorial process. 2024. Available from: https:\/\/www.elsevier.com\/about\/policies-and-standards\/"},{"key":"pcbi.1014776.ref027","author":"T Sharpton","year":"2026"}],"container-title":["PLOS Computational Biology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1014776","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,16]],"date-time":"2026-09-16T17:34:05Z","timestamp":1789580045000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1014776"}},"subtitle":[],"editor":[{"given":"Francis","family":"Ouellette","sequence":"first","affiliation":[{"name":"Montreal","place":["CANADA"]}],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,9,16]]},"references-count":27,"journal-issue":{"issue":"9","published-online":{"date-parts":[[2026,9,16]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pcbi.1014776","relation":{},"ISSN":["1553-7358"],"issn-type":[{"value":"1553-7358","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,16]]}}}