Understanding the GeneRank Model

GeneRank is a new mathematical model in Bioinformatics, that is used in generating prioritized gene lists by exploiting biological background information. It is a modification of PageRank in the popular search engine Google. Using the matrix decomposition technique we find in this paper that there two dominant portions in the output of GeneRank. One is the information from the microarray data, and the other is the previous knowledge about the biological function of genes. The solution of the GeneRank can be approximated well by the linear combination of the two parts. Numerical results show the significance of the dominant parts. Thus the results of this paper give an understanding of GeneRank model.