Roma Nagle Rose Hills
Development of an RNA 3D Prediction Machine Learning Model
Structure determines function. This ground truth drives the exponential progress being made in biology. By understanding a molecule’s structure, scientists can harness its function for drug discovery, genetics, or even studying evolution. However, determining a molecule’s structure in the lab is not easy. Even with advancements such as cryoEM, there has been a significant push to computationally predict structures instead. This motivation is at the heart of my research in the Cate Lab. Can we use machine learning to predict the 3D structure of an RNA molecule from just its primary sequence? Our model will deviate from current approaches by relying on untapped structural homolog data. In other words, we will be relying on families of sequences with similar structures. We hope this allows for an increase in accuracy, a large bottleneck for the current standard of RNA prediction models. Ultimately, we plan on entering our model in the Critical Assessment of Structure Prediction (CASP) in 2024, a worldwide competition for the best RNA prediction method.
Message To Sponsor
Thank you for providing me with the opportunity to continue the research I love doing this summer. I'm very excited to learn more about the intersection of machine learning and structural biology, all while becoming a better researcher, data scientist, and student. Your contribution is part of the many reasons I will never forget my time in undergrad at UC Berkeley.