Skip to main content
Fig. 3 | Genome Biology

Fig. 3

From: Modelling local gene networks increases power to detect trans-acting genetic effects on gene expression

Fig. 3

Application of alternative eQTL association methods to the mouse dataset. af The number of detected trans eQTLs stratified by the number of regulated genes (hotspot size). Results for alternative significance thresholds are colour coded. While PC-based approaches increased power to detect variants that regulate a small number of genes (c, d), GNet-LMM (g, f) maintained a profile of eQTL hotspots that was similar to the results from an unadjusted analysis (LMM) (b). g Total number of trans associations retrieved by alternative methods as a function of the significance threshold. h Gene enrichment analysis of trans eQTLs using Reactome pathways. Solid lines show the proportion of Reactome-consistent trans eQTLs as a function of the rank of individual associations sorted by their significance level (“Materials and methods”). Dashed lines denote the corresponding p value threshold. eQTLs identified by GNet-LMM were markedly more enriched for Reactome pathways than eQTLs retrieved by alternative methods

Back to article page