GENE EXPRESSION DATA MINING FOR PHYSIOLOGICAL DISCOVERY

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. The main aim of this project is to develop novel gene array data analysis methods for the following problems which are fundamental and which are common to and shared by various aspects of the biomedical medical motivations described above. The main technical objectives of this project are as follows: Design and develop novel noise reduction, dimension reduction and data normalization methodologies, such that the transformed numerical indicators of gene expression intensity are comparable in terms of probability and gene function (Year-1). These methods will be applied to the gene array data from all three biomedical research projects described above. Design and develop fast and accurate algorithms for both supervised and unsupervised classification of comparative and time-series gene expression data (Year-1 and Year-2). Applicability will be to all three projects. Investigate and develop performance indices for the evaluation of the strength of data mining methodologies for fast gene expression profiling in large biological datasets, such as those provided by our collaborators, with a driving principle of supporting diagnostic discovery (Year-2 and Year-3). Applicability will be to all three projects. Apply and extend the developed algorithms on experimental datasets provided by our collaborators (Drs Lukiw, Hill and Cohen), and providing new computational and biological insight into their data, by applying novel data preparation and mining technologies (Year-2 and beyond). Using functional information from ontological databases and domain knowledge experts, to rediscover and refine discovered clusters, to be significant in terms of their coherent molecular-functions, and their participation in biological-processes and cellular-components (Year-3 and beyond). Applicability will be to all three projects and all other available data from INBRE researchers.