UCSD - University of California - San Diego

11/02/2026 | Press release | Distributed by Public on 11/02/2026 10:29

New $15M Department of Energy Genesis Award Will Use AI to Decode the Shape of RNA

Published Date

November 02, 2026

Article Content

Growth of the U.S. bioeconomy is vital to American prosperity and energy independence, and its future relies on the ability to engineer microbes and crops with predictable, desirable traits. Current models do not reliably predict how genetic modifications will alter those traits, and one major reason is that they omit the three-dimensional shape of ribonucleic acid (RNA).

A new three-year, $15 million award from the Department of Energy (DOE) will support University of California San Diego Professor of Biochemistry and Molecular Biophysics Navtej Toor in his effort to expand the library of known RNA structures and use it to train artificial intelligence (AI) models that predict the three-dimensional shape of RNA from its sequence alone.

The project, titled "Decoding the RNA Structurome to Secure AI Advantage for the Bioeconomy," is one of 12 Phase II Genesis Mission awards, totaling $159 million, announced by the DOE on October 8.

"The United States has a long tradition of leadership in groundbreaking science, and UC San Diego is proud to be part of that continuing effort," said Chancellor Pradeep K. Khosla. "Navtej Toor's research will significantly expand our understanding of how RNA functions within living cells and drive advances in biofuels, health care and the broader biochemistry of life. It also exemplifies our commitment to supporting bold, curiosity-driven research that creates impact and solves challenges across communities, industry and society."

RNA in the spotlight

Professor of Biochemistry and Molecular Biophysics Navtej Toor.

RNA is a single-stranded nucleic acid present in all living cells. While DNA has a simple double-helical structure, RNA can fold into a wide variety of complex structures. RNA is essential for protein synthesis, catalyzing chemical reactions and carrying genetic information, and its three-dimensional shape determines how it performs each of these roles.

Yet we do not know the structures of most RNA molecules. Artificial intelligence, which has already transformed the study of proteins, stands to accelerate the process. In fact, the 2024 Nobel Prize in Chemistry was given to the creators of AlphaFold - a software tool that can accurately predict protein structures after training on a dataset of 200,000 experimentally determined proteins. A similar revolution awaits RNA, but it first requires a large dataset of experimentally determined RNA structures on which to train. Fewer than 200 protein-free RNA structures have been determined over the past 50 years.

Toor's project will increase, by orders of magnitude, the speed at which RNA structures are solved and help supply the data that AI needs. "A data shortage, not computing power, blocks accurate AI prediction of RNA shapes and limits the design of organisms for energy and biomanufacturing," he said. "This project closes the RNA data gap through paired advances in experimental measurement and in AI algorithm design."

RNA forms complex three-dimensional structures that regulate gene expression. (cr: Toor lab / UC San Diego)

Determining the 3D structure of RNA is notoriously difficult. Crystallography can be used efficiently to solve protein structures, but it can take years to solve a protein-free RNA structure. Toor focuses on protein-free RNA structures because they reveal the shape RNA adopts without being obscured by the protein. This is the information AI models need to predict structure from sequence alone. Toor's lab holds the record for the largest protein-free RNA structure determined by X-ray crystallography, a project that took five years to complete.

Using cryo-electron microscopy (cryo-EM) to image RNA can also be tricky. Cryo-EM shoots electrons through a frozen sample, but protein-free RNA molecules are often too small and flexible to image clearly. The Toor lab found a way to stabilize the RNA by inventing a scaffold technology. With this scaffold, the lab has reduced the time to determine an RNA structure to only two weeks, compared with the years that crystallography can require. In the past year alone, the lab has determined 60 protein-free RNA structures, nearly a third of the field's 50-year total.

Toor and his collaborators will determine RNA structures from select DOE model organisms, including two photosynthetic microbes that make fuels and chemicals from carbon dioxide, and the bioenergy crops sorghum and poplar. DOE model organisms are chosen specifically to advance the missions set forth by the DOE Office of Science, including bioenergy.

Toor points out that this project is as much an AI innovation as it is an experimental one. "Algorithm development runs in a closed loop with the experiments: the models flag the RNA molecules they predict least confidently, those molecules are imaged next, and retraining on the results advances the AI throughout the project," he said.

His goal is to determine 1,000 protein-free RNA structures by the end of the project - five times as many as the entire field has produced in the past 50 years.

"Our goal is an online program that accepts any RNA sequence and returns an accurate three-dimensional structure," Toor said. "Eventually, the program could be applied to human RNA. In the long term, a pharmaceutical company developing a drug against a specific region of an RNA could obtain the structure of that target immediately and use it to guide drug design."

UC San Diego leads the project with co-investigators Rhiju Das of Stanford University, Wah Chiu of SLAC National Accelerator Laboratory and Daniel Haack of A-Form Solutions, Inc. Additional partners are the University of Washington, the Howard Hughes Medical Institute's Janelia Research Campus, Google and NVIDIA.

Competing interest disclosure: Toor is a co-founder of A-Form Solutions, Inc., which may potentially benefit financially from the research results.

Learn more about research and education at UC San Diego in: Artificial Intelligence

Read more news about: Physical Sciences, Artificial Intelligence, Health Innovation

UCSD - University of California - San Diego published this content on November 02, 2026, and is solely responsible for the information contained herein. Distributed via Public Technologies (PUBT), unedited and unaltered, on November 02, 2026 at 16:29 UTC. If you believe the information included in the content is inaccurate or outdated and requires editing or removal, please contact us at [email protected]