Shared genetic architecture in autoimmune disease - preliminary analysis

Diseases that have different underlying genetic risk component(s) may share similar phenotypes. Traditionally, disease classifications have focused on characterizing diseases based on sets of related phenotypes. As an example, type I diabetes and type II diabetes are both classified as a type of diabetes based on patients having high blood sugar over long periods of time. However, as our understanding of genetic contributions to disease susceptibility and progression evolves, we start noticing that diseases with similar symptoms may have completely different causes. For example, type II diabetes is due to insulin resistance while type I diabetes is caused by immune cells attacking insulin producing cells. As genetic data becomes more highly available, it becomes possible to classify diseases based on their genetic causative drivers instead of their phenotypes. In this study, we have (1) explored the relationship between 10 autoimmune diseases along with type II diabetes based on their genetic susceptibility information and compared such classifications to existing disease categorizations based on disease symptoms/phenotypes from Human Phenotype Ontology, NCI-thesaurus and the Disease Ontology, and (2) developed automated scripts to compute similarities and cluster diseases based on the specified criteria. Categorization based on genetic susceptibility can help identify diseases that share similar drug targets and benefit from similar diagnosis technologies. We hope to further develop our system to apply it to more disease categories.