Figure 3

Comparison of normalisation strategy. Relative expression analyses using either a single reference gene with low stability between wild-type (WT, open bars) and R6/2 (filled bars) cortex such as Actb, a single reference gene with high stability between the two genotypes such as calnexin or the geometric mean of three highly stable reference genes. P-values above the R6/2 bars show that while a single, stably expressed gene (p = 0.0016) is a better calibrator than a gene with low expression stability between the two genotypes (p = 0.0926), calculating the geometric mean of three stably expressed genes is a far superior normalisation strategy (p = 0.0006). Error bars are S.E.M (n = 6).