Hierarchical breakdown of RNA structure prediction in CASP16: from reliable local helices to speculative multimer assembly.

Hierarchical analysis reveals that RNA structure prediction accuracy declines from local helices to multimer assembly, with the highest errors in interface and tertiary interactions.

• Why it matters: Accurate RNA structure prediction, especially for multimers, is crucial for understanding RNA function and designing therapeutics, yet remains a significant challenge in the field.
• What they did: Using CASP16 submissions, particularly the top-ranked LCBio models, the study examined prediction performance across structural levels, identifying key failure modes through hierarchical analysis and case studies.
• The result: Findings highlight that improving monomer accuracy, interface modeling, and model selection is essential for advancing reliable RNA-RNA multimer predictions, guiding future methodological improvements.