RNASSTR: new Rfam-based RNA secondary structure dataset with structure-aware train/test splits

chaitjo · x · 2026-09-05

A team from UC Berkeley's Innovative Genomics Institute (Doudna and Cate labs) released RNASSTR, a larger RNA secondary structure corpus built on Rfam with structure-aware train/test splits. RNA structural data has long lagged behind protein data — the key to AlphaFold2's success — and this dataset, published in the journal RNA, aims to ease that training bottleneck for AI-based RNA structure prediction.

Original post →

More from Research

Research channel →