Deep learning coupled with scalable domain-specific structural validation expands RNA virus discovery from metatranscriptomes
Gaoyang Luo, Zelin Zang, Ling Yuan, Jingbo Zhou, Ao Dong, Yufei Huang, Stan Z Li, Feng Ju · 2026
Read Deep learning coupled with scalable domain-specific structural validation expands RNA virus discovery from metatranscriptomes by Gaoyang Luo, Zelin Zang, Ling Yuan, Jingbo Zhou, Ao Dong, Yufei Huang, Stan Z Li, Feng Ju (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13677681.txt.
Description
RNA viruses hide in environmental samples in wildly varied forms, from giant fused proteins to broken fragments, which defeats many search tools. Gaoyang Luo and co-authors present Rider, a two-stage hunter. First, a compact protein language model scans broadly for faint viral signals. Then a sliding-window structure prediction stage checks the shapes of candidate regions against a curated library of the virus-copying enzyme's core domains.
Across more than 10,000 samples from many ecosystems, Rider matched the sensitivity of the leading tool, recovering over 99 percent of known RNA viruses. It also found thousands of deeply divergent sequences and families that other methods miss. In a human gut-disease sample set it agreed with rival tools while reaching further. The authors offer it as a scalable route to mapping the world RNA virome, in 2026.
Catalog record
| CREATOR | Gaoyang Luo, Zelin Zang, Ling Yuan, Jingbo Zhou, Ao Dong, Yufei Huang, Stan Z Li, Feng Ju |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 113 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |