With this award the Chemistry of Life Processes Program in the Chemistry Division supports Drs. Mary Jo Ondrechen and Penny J. Beuning from Northeastern University to develop, implement, and verify a novel method for predicting the biochemical function of proteins. This project utilizes computational chemistry tools in an innovative chemical-properties-based method to discover the biochemical function of protein structures. It begins with the premise that local structural features at the active site constitute better predictors of biochemical function than does the overall fold. First it is shown that sets of structurally similar proteins may be sorted into subgroups according to biochemical function, using the active site residues predicted from computed chemical properties, structurally aligning them, and scoring them according to chemical similarity. Thus for each functional subgroup, a characteristic chemical signature are defined. Next, structural genomics proteins are analyzed and annotated, using their calculated chemical signatures and seeking matches with the chemical signatures of the previously characterized proteins, and misannotated structural genomics proteins are identified. For selected cases of structural genomics proteins that have been determined to be misannotated, or for which biochemical function is newly predicted, the predicted functions are verified experimentally by kinetics assays and binding studies.
Genome sequencing and structural genomics efforts over the past 15 years have provided a wealth of new information about thousands of proteins. While genome research holds tremendous promise for major future benefits to society, a key step toward the realization of this potential is the ability to determine the function of the thousands of protein structures whose biochemical functions are currently unknown or uncertain; this project addresses that current challenge. Greater understanding of protein function has many potential benefits in fields outside of chemistry, including agriculture, biotechnology, medicine, biofuels, and "green" industrial processes. A better understanding of the local structural, electrostatic, and chemical patterns that govern how proteins work so selectively and efficiently will open the door to novel commercial applications. Such basic knowledge to drive novel high-tech commercial application is important to the future growth of our regional and national economies. This project also will develop a knowledge base and tools that will enable other discoveries and applications. In this project a racially diverse group of both undergraduate and doctoral students are trained in methods of computational chemistry, computational biology, and biochemistry. Experience in these areas is a vital ingredient for United States competitiveness in the global economy. Individuals with such knowledge and training are in high demand in both academia and industry.