Power calculations determine the optimal Bdnf assays for use as preclnical assessment tools. (A) Shows the relative expression level of each Bdnf promoter-specific transcript expressed as a ratio of R6/2: wild-type from a pre-symptomatic time point (4 weeks, 4 w) to early symptomatic (8 w) and late symptomatic time points (12 and 15 w). Therefore, each bar represents the relative amount of down-regulation of each transcript in R6/2 mouse cortex compared to wild-type. Significant differences between R6/2 and wild-type mice for a specific transcript at the ages indicated are represented with asterisks below the relevant bar. * p < 0.05, ** p < 0.01, *** p < 0.001. Promoter-specific transcript nomenclature is according to Liu et al., 2006 . We performed power calculations (shown is 80% power of detection of improvement at p > 0.05) as previously described  for BDNF promoters, with promoters I (B), IIa (C), V (D) and VIII (E) being the best powered to detect potential modulation of Bdnf expression levels by genetic or pharmacological approaches (data not shown for remaining promoters). The dotted line illustrates the number of mice needed in order to be able to detect a 30% improvement with 80% confidence.
Benn et al. Molecular Neurodegeneration 2008 3:17 doi:10.1186/1750-1326-3-17