De novo design of RNA pseudoknots with deep learning
- Jill Townley
- Wipapat Kladwang
- David Baker
- Hamish M. Blair
- Christian A. Choe
- Gina El Nesr
- Andrew Favor
- Eli Fisker
- Daniel B. Haack
- Shujun He
- Jason Hingey
- Po-Ssu Huang
- Rui Huang
- Chaitanya K. Joshi
- Thomas Karagianes
- Andrew Kubaney
- Pietro Liò
- Adamo Mancino
- Jonathan Romano
- Boris Rudolfs
- Nicholas Spellmon
- Navtej Toor
- Jigyasa Verma
- Vivian Wu
- Zhiheng Yu
- Eterna Participants
- Rhiju Das
2026-08-27
RNA design has been hindered by the limited accuracy of three-dimensional (3D) structure prediction. In this study, we show that intricate RNA structures can be generated with current deep learning tools through accurate de novo design of pseudoknot secondary structures. In an Eterna competition involving 57 pseudoknots, generative artificial intelligence (AI) methods matched experienced human designers in solving most blind challenges, evaluated by single nucleotide–resolution chemical mapping, compensatory mutagenesis, and cryo–electron microscopy. AI-generated molecules with accurate secondary structures formed well-ordered 3D folds stabilized by noncanonical tertiary interactions not modeled during design. Success was guided by an RNet foundation model trained on prior chemical mapping data, suggesting that some difficult RNA design tasks may be tractable without first solving RNA 3D structure prediction.