Stanford Team Releases Minerva, a Genome Language Model for Biological Discovery
David Li and Garyk Brixi from Stanford's Brian Hie lab, in collaboration with the Michael Fischbach lab, have released Minerva, a framework for scientific data mining powered by genomic language models; the preprint is now public. The work demonstrates a path to systematically discovering genetic elements with AI and accelerating biological design—another case of language models reaching into fundamental biology.
Confirmed
- Minerva is a method for biological discovery using genomic language models, developed by David Li and Garyk Brixi under the supervision of Brian Hie, with gratitude to advisor Michael Grinstein and in collaboration with the Michael Fischbach lab.
- The preprint is publicly available for the research community to read and reuse.
- Core finding: the UG27 reverse transcriptase system encodes diverse arrays of ncRNAs with shared structure, each ncRNA templating a short DNA hairpin.
Why it matters
- According to Brian Hie, the study shows that AI can not only accelerate biological design but also systematically discover new genetic elements and reverse transcriptase mechanisms, providing a reusable framework for genomic data mining.
2026-09-23 ~ 2026-09-23 · 7 related posts
Primary sources
- [source] Genome language models uncover new class of reverse-transcriptase mechanisms — BrianHie · 2026-09-23
- [source] Minerva paper details: ncRNA arrays template short DNA hairpins — BrianHie · 2026-09-23
- [source] Minerva: Hie lab's data-mining framework for scientific discovery — BrianHie · 2026-09-23
- Minerva framework preprint is out, says Brian Hie — BrianHie · 2026-09-23
- Minerva: genome language models uncover diverse structured ncRNA arrays in biology — BrianHie · 2026-09-23
- Minerva: genome language models uncover co-evolutionary systems across prokaryotes — BrianHie · 2026-09-23
1 near-duplicate retellings: BrianHie