Computational prediction and analysis of protein-RNA interfaces

UnrestrictedProtein-RNA complexes play a central role in many biological processes, including translation, mRNA stabilization, ribosome formation, and viral packaging. The investigation of short nucleic acid polymers (aptamers) and their interactions provides a greater understanding of biological phenomena. By virtue of their molecular characteristics such as high specificity and affinity, nucleic acid aptamers are amenable to being designed for therapeutic and diagnostic applications. The goals of this thesis are to create a non-redundant database of protein-RNA interactions, to generate a mechanism for evaluating the success of computationally predicted nucleic acid aptamers, and implement a method of distilling the evolution of a protein-RNA interface during a molecular dynamics simulation down to only the unique conformations. We hypothesized that this may be achieved by creating an algorithm which would reliably assess the similarity of one protein-RNA interface to another protein-RNA interface. When given a list of protein-RNA complexes or in the case of a molecular dynamics simulation, a dry coordinate file, the algorithm correctly locates each protein-RNA interface, examines the interface contacts, geometry and motifs, and performs an all-to-all comparison