CDS&E: Computational Design of Peptide-Based Biorecognition Elements

North Carolina State University, Raleigh NC

Investigators

Abstract

Peptides are used in numerous applications: from discovery of biochemical threats to the public, to detection of molecular markers of disease, to purification of therapeutics. The goal of this project is to develop algorithms that could be used to design peptides to bind to a single target protein. To help test the algorithms and tailor them to the needs of a variety of researchers, the team will partner with scientists and engineers to perform experiments to see if the designed peptides work. The projects include designing peptides for use in separations, counteracting bacterial infections, and early detection of cancer. The methods developed should provide fast and effective guidance to industry and academia on which peptides to choose for which applications. The PI will continue to promote inclusion of underserved groups in science by meeting with women graduate and undergraduate students. A video presentation about peptide design aimed at general audiences will be created and posted on YouTube. A fun iPad app/game/puzzle to illustrate the challenges associated with binding between peptides and proteins in drug development will be crafted for use by high school and college teachers. This Computational and Data-Enabled Science and Engineering (CDS&E) project will provide new software and data analysis methods to accelerate discovery in a variety of engineering and bioscience disciplines that have peptides as their common thread. The goal is to develop user-friendly software that could be used to design short flexible linear peptides to bind to a single target protein. The algorithm will be tailored to modern problems requiring peptides: that are designed de novo (without reference peptide), that mimic antibody tips, that bind to surfaces, that bind to one protein but not another, that bind to two similar proteins, or that link two different proteins. To help test the code and tailor it to the needs of a variety of researchers, the team has partnered with four members of the engineering/science community, each of whom has conceived a “demo project” --- the Hall lab will design key peptides needed in the partner’s research, and the partner will do experiments to see if the designed peptides work. The projects include designing peptides: to attach to resins in chromatography, act as molecular glues to precipitate antibodies, deactivate toxins in bacterial infection, bind membrane-spanning proteins on exosomes, and mimic antibodies that bind to closely related variants. The software will include a machine-learning component to predict peptide affinity based on the amino acid sequence and the structure of the peptide-protein complex. This work should have a transformative long-term impact on researchers in a variety of fields who need specialized peptides for their applications but are not modelers, or not comfortable with modeling. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.