Undergraduate Thesis and Graduation with Distinction Defense
Computational Design of RNA with Perturbations
||Monday, April 17, 2017
||10:00am - 11:00am
||D106 LSRC, Duke
Like proteins, RNA molecules are responsible a variety of essential functions within the cell. They are flexible and complex molecules that display intricate secondary and tertiary structures, and serve many roles in addition to protein translation. They exhibit catalytic activity, either alone (ribozymes) or as part of complexes with proteins (ribonucleoproteins). The goal of this work is to extend computational protein design algorithms so that they can be applied to systems involving RNA as well. Computational protein design is a field of research that seeks to determine the sequences of amino acids that will produce the lowest energies for a specific backbone to create proteins with desired shapes and functions. Assuming a fixed backbone and that sidechains take on conformations from a discrete, experimentally observed set of rotamers, the DEE/A* algorithm will provably find the global minimum-energy conformation (GMEC). Similar methods can be used to determine the sequence of ribonucleic acid residues that will minimize the energy of an RNA structure by introducing RNA “rotamers” that sample the experimentally observed chi angles. Furthermore, a novel protein design algorithm (DEEPer) developed in the Donald Lab allows for small perturbations in the protein backbone, such as backrubs and shears, while still provably finding the GMEC. This work proposes a perturbation specific to the RNA backbone to allow the nitrogenous bases to shrug in a similar fashion to the protein backrub. Several designs that incorporate these new RNA-specific techniques will be examined.
Hosted by: Bruce Donald