Stochastic variation of transcript abundance in C57BL/6J mice

BMC Genomics

Table 1 Variability of transcript abundance

		Adipose	Heart	Kidney	Liver	Total
(A)	Expressed Genes	11204(49%)	10069(44%)	10116(44%)	8919(39%)	12657(55%)
(B)	Variable Genes
	α= 0.05	3923(17%)	2654(12%)	2273(10%)	2125(9%)	6932(30%)
	α= 0.01	3299(14%)	2117(9%)	1547(7%)	1539(7%)	5800(25%)
	α= 1e-4	2352(10%)	1250(5%)	847(4%)	813(4%)	3939(17%)
(C)	Between-Mouse Variation
	FWER (p < 0.05)	0(0%)	0(0%)	12(0%)	40(0%)	46(0%)
	FDR (p < 0.10)	0(0%)	0(0%)	369(2%)	2520(11%)	2674(12%)
	1 - π ₀	0%	0%	1%	18%
(D)	Maximal Fold Change
	> 1.5	2103(9%)	509(2%)	320(1%)	538(2%)	2690(12%)
	> 2.0	600(3%)	142(1%)	91(0%)	809(4%)	809(4%)
	> 3.0	218(1%)	50(0%)	28(0%)	36(0%)	292(1%)

Shown in this table are: The number of expressed genes out of the 22869 annotated genes on the array (A). The number of variable genes based on the 0.05, 0.01 and 0.0001 tails of the scaled χ₂(23) distribution (B). The number of genes with significant between-mouse variance, based on the F_s test with family-wise adjusted error (p < 0.05, Sidak step-down method [5]); false discovery rate (p < 0.10, Benjamini-Hochberg method [6]); and the estimated proportion of differentially expressed genes (1 - π₀, using the q-value method [7]) (C). The number of genes with large maximal fold-change between mice (D). All numbers are given as percentages of the annotated genes in parentheses.

ISSN: 1471-2164