We are writing a modelling program that will be used for biasing combinatorial libraries for finding RNA-binding peptides. Currently the complexity limit of our in-vivo combinatorial libraries is 109, which translates into the ability to randomize 5-6 residues with 20 amino acids. We wish to randomize more positions, possibly as many as 20 by using a computer model to do some of the combinatorics, resulting in a more limited set of amino acids. By limiting the search space for each position more positions can be randomized. We will first write the modelling program in C which will take a binding site as input and output the optimal combination of amino acids docked into the binding site to maximize the hydrogen bonding interactions. We are testing this program with is the arginine-rich (ARM) peptides which rely largely on hydrogen bonding interactions. The program, in addition to requiring computational power, requires visualization and possibly energy minimization.
Showing the most recent 10 out of 508 publications