Figure 7
From: Mixture modeling of transcript abundance classes in natural populations
Strength of association between cis-SNPs and transcript abundance. Frequency histograms in bins of increasing order of magnitude of significance, with number of genes indicated on the y-axis. (a) The distribution of significance measures (negative logarithm of the P value) for the most strongly associated single nucleotide polymorphism (SNP) within 100 kilobases of each of the 818 bimodally expressed transcripts. (b) The same distribution for SNPs linked to a set of 835 randomly selected transcripts. Note the excess of outliers in the bimodal sample. (c) The distribution of strongest associations for a typical permutation of SNPs against unlinked transcripts, clearly showing much reduced significance relative to those observed for linked SNPs. (d) The 'best possible' distribution of associations, assuming that a single SNP explains all of the observed bimodality of each transcript. Single dots in panels A and D represent outlier significance values.