Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences
PNAS 118(15), e2016239118, published 5 April 2021. Submission and acceptance from the article's own editorial line ("accepted by Editorial Board Member William H. Press December 16, 2020 (received for review August 6, 2020)"); publication date from the PMC citation metadata (5 April 2021).
- Issued by
- No issuing record yet. A missing record is a gap in coverage.
- Authors credited
- Zeming Lin · Alexander Rives
- Dates
- Published 2021-04-05 authorship evidence authorship evidence
- Exact source
- Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences Proceedings of the National Academy of Sciences (open access copy in PubMed Central) · published 2021-04-05 · Retrieved for the candidate pack on 2026-10-30 (SHA-256 5a365c9b9bca1d0cb6065822636366ef838a8192d09787229202ea2c5628b236). PMC8053943 copy of PNAS 118(15):e2016239118.
Appears in
No collection holds a record about this document yet.
Versions
No other versions of this document are recorded. The evidence applies to the source version listed above. Other versions, such as a preprint, working paper or journal publication, are not identified or substituted.
Issue and authorship records
- Author: Zeming LinZeming Lin is credited as a co-author of Rives et al., ‘Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences’ (PNAS, 2021).2021-04-05 · publication · Stated in source · Supported Inspect claim & evidence ↗Zeming Lin: person page →
- Author: Alexander RivesAlexander Rives is credited as the lead-listed author of ‘Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences’ (PNAS, 2021).2021-04-05 · publication · Stated in source · Supported Inspect claim & evidence ↗Alexander Rives: person page →
Documentary relations
Each relation is listed separately. Citing, crediting and reporting the use of a method are different acts, and none of them shows agreement or influence by itself.
Not recorded for this document: cites · cited by · explicitly credits · reports using a method from · method reported as used by · applies parts of · parts applied by · discusses · reports evaluation of. A missing record does not show that the relation does not exist.
What the document states
A limited selection of what this document states, each with its passage. It describes the text, not whether anyone acted on it.
- StatementUnsupervised training on 86 billion amino acids across 250 million sequences“we use unsupervised learning to train a deep contextual language model on 86 billion amino acids across 250 million protein sequences spanning evolutionary diversity”abstractInspect claim & evidence ↗
- StatementProtein language modelling framed as a step towards predictive and generative AI for biology“Protein language modeling at the scale of evolution is a logical step toward predictive and generative artificial intelligence for biology.”abstractInspect claim & evidence ↗