Understanding the causal pathways that link genotypes with gene expression and higher order phenotypes, such as disease susceptibility, has been a recent goal of systems genetic studies. Pairwise correlations can select genes and phenotypes that are associated with a common genetic locus; however, the directions of these links are more difficult to determine. We propose a method in which Bayesian networks are used to establish causal pathways from genotype to gene expression to phenotype and apply it to mice infected with Chlamydia psittaci.
Recombinant inbred BXD mice strains have been previously used to study the genetic differences that cause a much greater resistance to Chlamydia psittaci infection in C57BL/6J mice compared with DBA/2J mice . Gene expression levels and higher order phenotypes, such as pathogen load and survival, were collected for 44 BXD strains. Directed graphs linking SNPs  in the BXD genomes to this data were created with QTL mapping and the Bayesian network method. Figure 1 shows a selection of the resulting causal pathways; many of the genes in these pathways are known to be associated with immune response or GTPase activity, which has been linked to Chlamydia psittaci infection .
Figure 1. Causal pathways linking genotype SNPs to gene expression levels to phenotypes in BXD mice infected with Chlamydia psittaci
This work was support by NIH grant AI081050 and DOD grant W81XHW-05-01-0227.
Miyairi I, Tatireddigari VRRA, Mahdi OS, Rose LA, Belland RJ, Lu L, Williams RW, Byrne GI: The p47 GTPases Iigp2 and Irgb10 regulate innate immunity and inflammation to murine Chlamydia psittaci infection.