Innovative Computational Methods for Protein Structure Prediction, Drug Discovery, and Therapeutic Design

The protein folding problem – i.e. predicting the three-dimensional structure of a protein from its amino acid sequence alone – is an unsolved problem in computational biology. I will present a novel computational approach to this problem that assembles protein tertiary structure from predicted secondary structure elements. I will demonstrate how this algorithm can be combined with limited EPR- and NMR-spectroscopic data to determine the structure of membrane proteins, an important class of proteins targeted by therapeutics. Structure- and Ligand-based computer aided drug discovery (CADD) algorithms are leveraged to develop small molecule therapeutics that bind to proteins. I will introduce ROSETTALIGAND, an algorithm that allows protein-small molecule docking with full protein and ligand flexibility. I will illustrate the application of machine learning algorithms for ligand-based CADD, specifically the discovery of allosteric modulators of human brain receptors for the treatment of schizophrenia and other neurological diseases.