System to predict novel genetic disease associations

Reese, Martin

Abstract

Rapid advances in biotechnology and clinical studies have produced an information bottleneck in data processing that has hindered application of such critical scientific data for use in medicine and disease diagnosis. Omicia will develop a novel genetics based informatics infrastructure to overcome this bottleneck that can predict the functional outcome of gene mutations and their relationship to human disease by integrating diverse knowledge bases in a consistent and structured manner. Implementation of the technology will reduce the cost and enhance the effectiveness of genetic association studies, population profiling analyses and pharmacogenomic applications, as well as potentially identify novel gene targets for disease. Our technology will mine classification systems for large number of genes and create statistical associations as well as integrate ontologies to allow for reliable inferences about functions of genes and their relation to disease phenotypes. Importantly, the predictive capabilities of the technology will assign functions to orphan genes and genetic markers for which very little functional information is available. The IT system will consist of a classification of genes into families using shared attributes in a number of well-structured domains to create a gene inference system (GIS) by comparing gene ontologies in GO with disease ontologies in MeSH. It will also use the well-annotated HGMD database to infer novel disease-related function to unannotated polymorphisms such as those in dbSNP. These studies are designed to demonstrate the feasibility of the technology so that it can be validated through future genetic association studies in a Phase II SBIR.