An Ontology-Based Approach for Computational Phenomics: Application to Autism Spectrum Disorder

Much still remains unknown about causal pathways between genes and complex traits in common disorders, such as cardiac disorders and mental illness. Success in genetic studies of such heterogeneous conditions may depend upon having a well-defined, standardized set of phenotypes that can support analyses of homogeneous subgroups (‘forward genetics’) or groupings based on available genetic marker data (‘reverse genetics’).  Phenomics—the systematic cataloging of phenotype terms on a genome-wide scale—is still emerging as a scientific field.  A critical limitation to its growth is the lack of informatics tools to characterize, manage, and analyze phenotypes.  In this talk, I will discuss research on a knowledge-based approach for phenomics, which includes the use of Semantic Web ontology and rule languages to encode and organize phenotypes and the visualization of this knowledge to inspect relationships among phenotypes.  I will discuss the results of applying our approach to autism spectrum disorder in our collaborative work with NIH’s National Database for Autism Research.