De novo design of RNA pseudoknots with deep learning
Authors:
- Wipapat KladwangWKWipapat Kladwang
- David BakerDBDavid Baker
- Hamish M. BlairHBHamish M. Blair
- Christian A. ChoeCCChristian A. Choe
- Gina El NesrGNGina El Nesr
- Andrew FavorAFAndrew Favor
- Eli FiskerEFEli Fisker
- Daniel B. HaackDHDaniel B. Haack
- Shujun HeSHShujun He
- Jason HingeyJHJason Hingey
- Po-Ssu HuangPHPo-Ssu Huang
- 27 August 2026
- journal article
- research article
- Published by American Association for the Advancement of Science (AAAS) in Science
- Vol. 393 (6814) , 931-937
Abstract
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.Keywords
Funding Information
- National Science Foundation (2330652)
- National Institutes of Health (R35 GM122579)
- National Institutes of Health (R35 GM141706)
- National Institutes of Health (R01 AI165433)
- National Institutes of Health (U19 AI181881)
- Howard Hughes Medical Institute
- Howard Hughes Medical Institute
- Qualcomm Innovation Fellowship
- A*STAR Singapore National Science Scholarship
This publication has 52 references indexed in Scilit:
- Assessment of Nucleic Acid Structure Prediction in CASP16Proteins: Structure, Function, and Bioinformatics, 2025
- Functional Relevance of CASP16 Nucleic Acid Predictions as Evaluated by Structure ProvidersProteins: Structure, Function, and Bioinformatics, 2025
- Computational De Novo Design of Group II Introns Yields Highly Active RibozymesChemBioChem, 2025
- Results of the Protein Engineering Tournament: An Open Science Benchmark for Protein Modeling and DesignProteins: Structure, Function, and Bioinformatics, 2025
- Untangling the pseudoknots of SARS-CoV-2: Insights into structural heterogeneity and plasticityCurrent Opinion in Structural Biology, 2024
- Assessment of three‐dimensional RNA structure prediction in CASP15Proteins: Structure, Function, and Bioinformatics, 2023
- Principles for Predicting RNA Secondary Structure Design DifficultyJournal of Molecular Biology, 2016
- The architectonics of programmable RNA and DNA nanostructuresCurrent Opinion in Structural Biology, 2006
- Encapsulating Streptomycin within a Small 40-mer RNAChemistry & Biology, 2003
- RNA pseudoknotsCurrent Opinion in Structural Biology, 1992