Two-stage ESM-2 screening plus structural validation uncovers divergent RNA viruses
bravo_abad · x · 2026-09-22
Luo and coauthors apply a "cheap screening, expensive validation" two-stage pipeline to find divergent RNA viruses in metatranscriptomic data—viruses whose sequences may barely resemble known ones.
- Stage one is deliberately broad and cheap: a compact ESM-2 protein language model plus a lightweight Transformer scans protein sequences for candidates resembling RNA-dependent RNA polymerases, the hallmark enzyme of RNA viruses.
- Stage two runs the expensive step only on candidates: 3D structure prediction compared against a curated library of catalytic polymerase domains. Structure stays conserved long after sequence similarity fades, catching divergent viruses.
The takeaway: when the search space is enormous, the expensive method should not examine everything—a model AI-for-Science pairing of deep learning screening and structural validation.
More from Research
- Microsoft's ShieldVLA Cuts VLA Safety Costs by 57% Using HJ Reachability — MicrosoftResearch · 2026-09-22
- UCSC's Prediction-Powered Smoothing Makes Disaggregated AI Evaluation More Precise — UCSantaCruz · 2026-09-22
- Rat-brain AI models land on AWS as biological computing goes mainstream — nordicinst · 2026-09-22
- OpenAI's unreleased model reportedly solved 100+ open math problems after 24 days of training — Confident_Salt_8108 · 2026-09-22
- Researcher praises review quality but questions whether peer review itself is worth the effort — moarbugs · 2026-09-22
- MIT's Buehler shows AI swarm self-organizing into hub-and-spoke scientific instrument — algo_diver · 2026-09-22