Genetical Genomics Analysis Software

DESCRIPTION (provided by applicant): Response to drug treatment is thought dependent upon genotype for many modern therapies. Knowledge of how each genotype responds to a particular therapy is bene?cial only in that one can identify portions of the population which cannot reap the benefits of said treatment. A better course of action is to identify not only which genotype responds, or not, to a particular therapy, but to identify which region of the genome is responding, or not, and how. We believe that this information will lead to new drug targets and better therapies that benefit a larger portion of the population. The goal of this proposal is to provide a suite of software tools for genetic and genomic scientists performing gene mapping experiments with genomic data as the response variable. These tools will ideally provide functionality for 1) detecting polymorphic regions of the genome that con- fer transcript expression differences, 2) identify polymorphic regions of the genome that impart expression differences in genes located elsewhere in the genome, and 3) detecting interactions between loci that may correspond to epistatic effects on transcription. Some software already exists to perform each of these tasks as distinct independent solutions. This proposal intends to produce an integrated solution, S+EQTL (S-PLUS for expression quantitative trait loci mapping), that utilizes the power of S-PLUS and both incorporates and extends the functionality of an exist- ing genetics suite. By providing scientists with an integrated set of tools for genomics experiments with a genetic component, more productive time can be spent interpreting the results rather than transforming data into different formats to be processed by multiple software analysis packages. This software should also address one of the most dif?cult aspects of genetical genomics exper- iments, the so called curse of dimensionality. As the genomics community continues gathering knowledge of transcripts in various organisms, the arrays that interrogate transcript abundance only grow larger in the number of transcript species included. In the absence of tools designed for this purpose, the research scientist is left with the option of either focusing on a narrow set of previously known genes or performing a grid-wise search on all genes in the array. The former is not interesting as these genes are likely well studied and may provide little novel insight. The latter is computationally demanding and may not be possible on the new, larger arrays. A recent publication presents a novel solution that may be enhanced to gain both power and scale using Bayesian methodology. Knowledge of how each genotype responds to a particular drug therapy is beneficial only in that one can identify portions of the population which cannot reap the benefits of said treatment. A better course of action is to identify not only which genotype responds, or not, to a particular therapy, but to identify which region of the genome is responding, or not, and how. We believe that the development of analytic tools for gene mapping experiments to identify this information will lead to new drug targets and better therapies that benefit a larger portion of the population.