Mapping and characterization of two relevance networks from SNP and gene levels

Variations of gene expression and DNA sequence are genetically associated. The goal of this study was to build genetic networks to map from SNPs to gene expressions and to characterize the two different kinds of networks. We employed mutual information to evaluate the strength of SNP-SNP and gene-gene associations based on SNPs identity by descent (IBD) data and differences of gene expressions. We applied the approach to one dataset of Genetics of Gene Expression in Humans, and discovered that both the SNP relevance network and the gene relevance network approximated the scale-free topology. We also found that 12.09% of SNP-SNP interactions matched 24.49% of gene-gene interactions, which was consistent with that of the previous studies. Finally, we identified 49 hub SNPs and 115 hub genes in their relevance networks, in which 27 hub SNPs were associated with 25 hub genes.