Recently, peptides have regained favor within drug discovery circles because there is mounting evidence that peptides, unlike small molecules and antibodies, can be used to significantly increase the space of druggable targets to include intracellular protein-protein interactions and agonists of class B GPCRs. In order to fill this biological gap, peptides must exhibit complex pharmacological behaviors, typically by binding to shallow and relatively featureless binding surfaces to modulate protein-protein interactions. The daunting task of designing novel peptides to modulate protein-protein interactions can be ameliorated by harnessing the vast amount of scientific information and knowledge of proteins that has accrued over the last two decades. Importantly, in an age variously referred to as the “computer” and “information” age, computation provides the ideal framework for harnessing and applying disparate sources of information to the problem of identifying therapeutic peptides to modulate protein-protein interactions. Here we describe a suite of computational, structure-based modules - implemented in our proprietary computational peptide drug discovery platform, CMDInventus - that can be used to solve a range of peptide lead discovery and optimization problems, from novel peptide binding site identification, to peptide structure prediction and high resolution protein-peptide complex prediction. Finally, we demonstrate a successful application of these tools to the de novo peptide design of antagonists of the CXCR4/SDF1 interaction.

Development, validation, and application of CMDInventus to enable structure-based peptide drug design and discovery

250th National Meeting (2015)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.