Kyle Roberts

In protein redesign one of the main challenges is the ability to evaluate many different conformations of the redesigned protein in order to determine the best structure. This problem is usually simplified by using a discrete set of side chain rotamers, and then using the dead-end elimination (DEE) algorithm to provably get rid of bad conformations. The Donald Lab has extended DEE to incorporate flexible side-chains (minDEE) and a flexible backbone (BD). These two extensions have since been used to redesign protein active sites and cores.

I am working to extend these algorithms further to design a protein’s ligand in order to design effective peptide inhibitors. My aim is to design a competitive peptide inhibitor for the CFTR/CAL system. The cystic fibrosis transmembrane conductance regulator (CFTR) is mutated in patients that have cystic fibrosis and the most common mutation of this protein is inefficiently synthesized and rapidly degraded. It has been shown that knock-down of CFTR Associated Ligand (CAL) can rescue the CFTR mutant function. The CAL PDZ domain has been shown to bind to the C-terminus of CFTR. The goal of my work is to design peptides that will efficiently competitively inhibit CAL and hopefully restore CFTR function in mutants.

Currently, my work has focused on extending the minDEE and BD algorithms to incorporate ligands of arbitrary amino acid length. Once this is complete, I will be able to use the algorithms to determine some potential peptide inhibitors. In addition, I will be able to compare how the algorithms rank peptides that have already been tested experimentally.