Prediction of the structure and function of proteins, based on their amino acid sequences, is the major challenge in the post genomic era. The major determinant of protein function is their three-dimensional structure, knowledge of which could be used to model protein-protein interactions and simulate complex signaling networks inside the cell. Knowledge of protein structure is also essential for rational structure-based drug design. Novel pathogen detection tools also could be developed based on knowledge of pathogenic protein structures. This project will develop an automated computer system, based on a machine-learning algorithm in combination with an extensive database of naturally occurring and computer-designed amino acid sequences, to predict the molecular structure of proteins from their amino acid sequences. Phase I will design and implement a preliminary prototype of a genetically applicable set of machine learning bioinformatics tools for detection of remote protein homologies and protein fold classification. Commercial Applications and Other Benefits as described by awardee: An automated computer system for protein structure prediction should greatly facilitate genome analyses as well as protein function prediction. The system also could facilitate the development of structure-based methods of pathogen identification and impact our ability to discover new treatments for human disorders.
