Talk 9
Presenter: Christian Deschepper
Wednesday, 2:00pm
M.P. Scott-Boyer, S. Picard and C.F. Deschepper Institut de recherches cliniques de Montréal (IRCM)
Using a panel of mouse AxB/BxA recombinant strains (RIS), we have previously identified on chr13 Lvm1, a QTL for cardiac left ventricular mass (LVM). After profiling gene expression in hearts from the same RIS and mapping expression QTLs (eQTLs), we found eight physically clustered cis-eQTLs and five trans-eQTLs matching the locus and profile of Lvm1 on chr13. By weighted gene co-expression network analysis, we found 49 gene co-expression modules. For the one correlating best with LVM, we detected a module QTL (mQTL) whose profile also matched that of Lvm1. Genes belonging to that module: 1) originated predominantly from the same chromosome that harbored the mQTL; 2) correlated with LVM in a fashion that was directly proportional to their connectivity; and 3) contained among its most connected genes the above 8 cis-eQTLs and 5 trans-eQTLs. Further analysis revealed that 21/49 modules showed evidence of being driven by a chromosome domain, since: 1) they showed linkage to one main mQTL located on one chromosome contributing a disproportionately high number of genes to the module, and 2) the genomic regions surrounding the peaks of these mQTLs showed enrichment in structural variants and polymorphic transposable elements. Thus, chromosome domains with particular physical properties and functional features constitute one principal component of certain co-expression modules, some of which have a QTL that matches that of a quantitative trait at the same location. The interconnectivity of genes in such modules greatly extends the number of candidate genes to consider as contributors to a phenotype within QTLs.