A Gene Retrieval System Using a Hereditary Disease Database

It becomes clear that most of genetic diseases and hereditary ones are caused by genetic variation and the abnormalities of gene expression. The genetic variation and the abnormalities are involved by one or more genes. In order to reveal the cause of these diseases, it is mandatory to enable us to retrieve the genes that are biologically related to such diseases. However, previous works considered that a pair of genes was assumed to be mutually related if an arbitrary pair of genes appears in the same literature. Therefore, the accuracy of retrieving related genes was low. In this paper, we discuss the way to retrieve biologically related genes by using OMIM database, in which functions of genes and their concerned genes are described for each genetic or hereditary diseases. We first generate a graph structure from OMIM database, in which diseases and genes are expressed as nodes and a gene node and a disease node are linked if these are appeared in the same OMIM database entry. Then, we obtain all of mutually and biologically related genes from the graph. This approach can improve the accuracy of retrieving related genes.