Student Research Symposium Program Portal: Submission #207

Abstract The p53 protein, or the “guardian of the genome,” is one of the body's biggest natural defenses against cancer. Therefore, understanding this protein could lead to big improvements in cancer research. But P53 is an incredibly dynamic protein, constantly moving and changing into different conformational states, because of this its extremely hard to model with traditional methods. Modeling the P53 protein using machine learning technology could be the thing to get accurate and insightful models. To research whether machine learning could create more accurate models than traditional methods, I will utilize AlphaFold to generate multiple models of the p53 protein and compare them to p53 models created using traditional methods that are found online in large databases such as ChimerX. If the models are as accurate or more accurate than traditional models, then I will use AlphaFold to create rarer conformational states of p53 and compare them to rare models created using traditional methods. If these models are as accurate or more accurate than traditional ones, I will finally attempt to create extremely rare or unknown conformational states. The real world significance of my research is not only a huge leap in cancer research but also a huge leap in complex protein modeling, which has a significant number of applications in many different fields.