DEV01260 Supplemental Data

Files in this Data Supplement:

• Supplemental Tables (PDF)
• Supplemental Figure 1 -

  Fig. S1. Identification of changes in gene expression in mice lacking Brn3a. Methods: Transcript levels were assayed in two independent sets of E13.5 trigeminal ganglia from Brn3a wild-type, heterozygous and knockout embryos. Of 12,422 transcripts represented on the U74Av2 chip, 4885 were present in both samples from at least one genotype. Two criteria were applied to detect significantly changed genes, the change-p value (Dp) and the fold change in hybridization signal. Transcripts that were significantly increased in Brn3a knockout ganglia have Dp <0.003, and transcripts that were significantly decreased have Dp >0.997. For the fold change criterion, the inclusion cutoff was set arbitrarily at a twofold change in the hybridization signal. The number of increased and decreased transcripts is greater here than in Tables 1 and 2 because ESTs are included in this analysis, but the inclusion criteria are the same.

  Comparison of the number of changed transcripts identified by the Dp or fold-change criteria in two independent experiments. When the Dp parameter was used alone to assess the number of changed transcripts, 108 transcripts were identified that were consistently increased in the absence of Brn3a (Dp <0.003), whilst 136 were reproducibly decreased (Dp >0.997). We then re-analyzed the data with the added criterion that to be counted as changed, genes must also exhibit at least a twofold change in relative expression between heterozygous and knockout ganglia. The combined application of the Dp and twofold change criteria reduced the number of transcripts changed in both experiments to 24 increased and 38 decreased genes.

  When each criterion is applied alone, there are two principal sources of discordance between the two experiments. Discordance in Dp occurs primarily because the overall hybridization signal strength was greater in Experiment 2, leading to a statistically significant Dp for a greater number of genes in that trial. Discordance in the fold change result usually occurs because of the application of an arbitrary cutoff of 2.0 for inclusion. The application of both criteria resulted in a high degree of concordance between the two trials, and transcripts that were significantly changed by both criteria in one experiment and not in the other usually fell just outside the inclusion criteria in discordant experiment. A list of neural genes that had significant Dp values but changed less than two fold, or were discordant between the two experiments, appears in Table S1.