The pseudoknot fold forms the core of most auto-catalytic RNAs, induces reading-frame shifts in viruses, and can act as a riboswitch. Predicting pseudoknots is difficult because the most popular RNA folding algorithms ignore them. More complicated and specialized pseudoknot algorithms neglect important structural constraints. Physical modeling is important for providing correct free energies for algorithms and for other insights. The Aalberts and Hodas model and statistical-mechanical theory predicts the probabilities of different pseudoknot folds, in good agreement with a database of known folds. The central focus of this RUI project is to extend and improve that statistical-mechanical theory to consider rarer types of pseudoknots, to study temperature and base-composition dependences of pseudoknots, to estimate theoretically the abundance of pseudoknot folds, and to test predictions against known structures. Theories are developed within the framework of polymer physics and statistical mechanics, numerical simulations employ RNA folding algorithms, and database analysis is done with the PI's own computer programs. An auxiliary goal is to develop and implement an efficient algorithm for predicting the most abundant class pseudoknots.
Williams College embraces scholarship by providing excellent facilities, promoting active faculty research, and running a remarkable summer research program with 170 students annually. Actively involving talented undergraduates in research projects is a priority because it provides inspiration and essential training for productive careers in the sciences by learning the tools they need to do cutting-edge research, and also how to present their work in refereed journals and through talks at conferences. Since 1998, the Principal Investigator has supervised nine undergraduate thesis projects with six refereed publications and two in process with undergraduate co-authors. The PI has also supervised nine other summer research students including two women. A number of colleagues and the PI are currently developing a new program in computational biology and genomics, and this project would help support training researchers in this growing field.
