Functional Annotation and Analysis of Genetic Variants

The process of functional annotation of human genetic variants is important for some significant biomedical tasks like understanding functional genomics and gene-disease associations. Moreover, genetic variants that lead to pathogenic mutations can have harmful consequences which may lead to certain medical conditions or diseases. This paper investigates and presents two methodologies utilizing the Gene Ontology (GO) for functional annotation of genetic variants (and human gene mutations): (1) using the concept of term enrichment from the biological process taxonomy in the gene ontology, and (2) using the concept of least common subsumer (LCS) within the gene ontology tree. The second methodology induces the most significant and accurate functions for a given set of variants having one common aspect, e.g. specific disease. This method is based on the structure of the directed acyclic graph of GO to induce and identify the most significant LCS's to be used for functionally annotating the variants under investigation. We applied the methods on a large sets of genetic variants causing mutations. Our method can determine important functions and with significance level of p<0.02, the outcomes and results of the method are biologically significant. We found that certain ontology annotations are more related to mutations through certain genes compared with normal bp functions (biological process terms) in GO; for example, one of these bp functions is {GO:0031325; positive regulation of cellular metabolic process}. The results are basically important and suggest that a mutation can be annotated with functions from the gene ontology, e.g. biological process aspect, just like the genes. This outcome may contribute into a more complete understanding of mutation gene-disease relationships, mutation pathogenicity, and understanding disease mechanisms.