Knowledge-Based Identification of Homogenous Structures In Gene Sets

The analysis of gene expression data plays a crucial role in disease diagnosis. However, such analysis is rather complex due to the high-dimensional gene space. This works shows that semantically related genes can be grouped using biological knowledge in the Gene Ontology (GO) database. The GO defines knowledge phrases called GO terms describing biological functions of genes. These knowledge phrases are typically used when discussing individual components of the ontologies. By exploiting the term frequency-inverse document frequency (tf-idf) as a distance measure between two genes, this distance is used in structure analysis on four gene sets causally associated with pain and the chroni-fication of pain, hearing loss, cancer, and drug addiction. The results show distinctive homogenous groups in each set of genes. This work outlines the successful application of information retrieval dissimilarity measure and structure analysis on gene sets using the knowledge base of the Gene Ontology. Furthermore, it enables in the future a separate analysis of gene subspaces.