Beyond the differences between the two association-based (HBCGM and GWAS) methods, the more extreme differences between association- and linkage-based methods were not factored into comparative modeling studies, which can produce misleading conclusions about the capabilities of murine GWAS. For example, one modeling study [3] used 723 loci affecting gene expression that were identified in linkage studies involving 2 inbred strains as the ‘gold standard’ for assessing the ability of GWAS involving 15 strains to identify causative loci. The poor correlation between the loci identified by the two different methods led them to incorrectly conclude that murine GWAS would not be productive. However, there is no reason to believe that genetic factors identified by analysis of gene expression differences across >10 inbred strains would be identical to those identified by linkage analyses involving two strains. For example, analysis of only two strains would present a myopic picture that would not reveal that there are three distinct levels of H2-Eα mRNA expression among the 10 strains that we analyzed [7] (Figure 2). H2-Eα expression levels varied by 268-fold across the 10 strains,
