The Proteomics data generated at the Proteomics Research Center is similarly being analyzed using sophisticated computer algorithms and software, much of it developed in house for specific aspects of each individual project. In particular, we are attempting to make use of an integrative analysis environment for interrelating information relating protein structure and function, protein expression, subcellular localization, and genotype/phenotype. The aim is to generate a cross platform proteomics data infrastructure that will both improve the quality and scientific content of our high-throughput MS-based proteomics efforts, and allow MS datasets to be integrated with DNA microarray, gene sequence and gene annotation databases. We have assembled a strong interdisciplinary team of bioinformatics researchers and computing engineers inspired to be at the forefront of the relevant disciplines.

For instance, in our efforts to extract meaning from protein expression profiles, we attempt to correlate protein abundance profiles with corresponding data describing the expression of cognate messenger RNA species. Clusters of coordinate turnover of proteins will be established using a derivative of the clustering algorithm developed to tackle microarray data. These algorithms uses quantitative data to identify pairs of gene products that behave similarly in a single experiment and then progressively adds other polypeptides to the initial nodes that serve to form clusters of co-regulated proteins. A cluster can provide a platform for probing the basis of cell regulation or function given an appropriate statistical threshold. Furthermore, one may identify and analyze sets of coregulated proteins for known and new properties, and assess if any of these elements (or variations thereof) contain information predictive of protein function or regulation. Follow up molecular genetic studies would then serve to substantiate the biological significance of this co-regulation.