The Research On The Shortest Metabolic Pathway Of KEGG

With the human genome project completion, people moving from the structural genomics to functional genomics research. Researchers have accumulated a great deal of biological knowledge, and have built many biologic databases. Using existing information resources, people can access data out of the work of retrieving data, They can concentrate on the analysis of data drawn from the databases.Bioinformatics makes the traditional biology more and more powerful, especially as the recent research focusing on the Microarray have the great power on the ability of solving the biological problem. With many gene expression data publicly available, the genetic research becomes more convenience. As human demands for the improvement of health and medical research in-depth, doing complex disease pathogenesis excavation is very necessary.This paper is based on the shortest path problem of KEGG pathway. First, based on dynamic programming algorithm we designed a metabolic pathway approximate matching algorithm, in the KEGG metabolism network tried to find the nearest best match. Then, we studied the gene expression profiles of genes mining principle and methods .Eventually, the research will focus on the core of mining the gene clusters concerning the shortest path problem of KEGG pathway. Using statistical methods we recognized the abnormal gene sets. Separately, the gene cluster concerning KEGG pathway consistent with the shortest path and random gene cluster using machine learning method to evaluate the effectiveness of their classification. The results showed that the shortest path gene cluster classification performs better than those random gene clusters. we finally choose a number of shortest paths gene clusters with the optimal effectiveness as the gene clusters concerning with metabolic pathways, eventually select a number of better performance gene clusters as the result of excavation.