Vaitea Opuu
CNRS researcher, BIOC, École Polytechnique
I design RNA molecules. My methods combine machine learning with physics-based models, such as molecular mechanics, to generate sequences that fold and function, and to measure how much information sequence data really holds.
Recent selected work
Nature Communications, 2025
Exploring the space of self-reproducing ribozymes using generative models
Generative models designed ribozymes that stay active up to 65 mutations away from the natural sequence and still copy themselves. Scaling up from the experimental success rate, at least 1039 such self-reproducing ribozymes may exist.
PLOS Computational Biology, 2026
A spectral framework for measuring diversity in multiple sequence alignments
The effective length Leff counts the genuinely independent information in a set of sequences. Natural RNA and protein families turn out far more constrained than their size suggests, and Leff anticipates how reliably their structures can be predicted.