WBPaper00046415:lithium_downregulated All data were analyzed using Stat View J 5.0, with all experimental data checked for assumptions of homogeneity of variance across manipulations using Bartletts test. Once the assumptions were satisfied, the data were analyzed by one-way analysis of variance followed by Dunnetts multiple comparison test. When homogeneity was not evident in the data, we used the nonparametric KruskalWallis test, followed by the MannWhitney U-test with a Bonferroni adjustment. Differences were considered significant at P < 0.05.
Genes suppressed following 24h exposure to lithium compounds (78uM LiCl and 375uM Li2CO3), according to custom DNA microarray.
WBPaper00035424:ASER_up Intensities of spot features annotated as Bad or Not Found in the .gpr files were set to 1 to be removed from further analysis, and all of the six processed .gpr data were converted to .mev file with TIGR ExpressConverter ver. 1.7. The .mev files were processed with TIGR MIDAS ver. 2.19 with parameters set as follows: one bad channel tolerance policy as generous, with both of channel flag checked and background unchecked. The data were normalized by lowess normalization with default settings and with block and slide SD regularization. Authors then calculated log2(ASER/ASEL) ratios for each gene on the microarray. For the two pairs of dye-swapped repeats, authors calculated the mean log2(ASER/ASEL) of each repeat, so that up to four log2(ASER/ASEL) values per spot were obtained. Authors then calculated the percentile rank for each gene. Each gene spot detected more than once (18 847 spots) were subjected to MannWhitneys U test to assess whether its percentile rank values are significantly higher compared to the rest of the genes detected in the same experiments. Resulting significance levels are shown by P-values. From the P-values, false discovery rate (FDR) was further calculated by the Benjamini and Hochberg method. Statistical analyses were done by using R software version 2.9.
Genes that showed higher expression level in ASER than in ASEL neuron by mRNA tagging.
WBPaper00035424:ASER_down Intensities of spot features annotated as Bad or Not Found in the .gpr files were set to 1 to be removed from further analysis, and all of the six processed .gpr data were converted to .mev file with TIGR ExpressConverter ver. 1.7. The .mev files were processed with TIGR MIDAS ver. 2.19 with parameters set as follows: one bad channel tolerance policy as generous, with both of channel flag checked and background unchecked. The data were normalized by lowess normalization with default settings and with block and slide SD regularization. Authors then calculated log2(ASER/ASEL) ratios for each gene on the microarray. For the two pairs of dye-swapped repeats, authors calculated the mean log2(ASER/ASEL) of each repeat, so that up to four log2(ASER/ASEL) values per spot were obtained. Authors then calculated the percentile rank for each gene. Each gene spot detected more than once (18 847 spots) were subjected to MannWhitneys U test to assess whether its percentile rank values are significantly higher compared to the rest of the genes detected in the same experiments. Resulting significance levels are shown by P-values. From the P-values, false discovery rate (FDR) was further calculated by the Benjamini and Hochberg method. Statistical analyses were done by using R software version 2.9.
Genes that showed lower expression level in ASER than in ASEL neuron by mRNA tagging.