A statistical approach to selecting and confirming validation targets in -omics experiments

BMC Bioinformatics

Table 2 Statistical validation analysis of data from six microarray experiments obtained from GEO

Study GSE #	# DE Genes	# Samples	Fraction of DE Genes	Cost	Cost
			Required for Statistical Validation	(Manual)	(Statistical)
GSE10245	6,742	58	3.57%	3.85 years	0.14 years
				$2.5e6	$8.8e4
GSE11492	333	8	72.37%	0.03 years	0.02 years
				$8.9e4	$6.4e4
GSE17913	739	79	32.61%	0.58 years	0.18 years
				$3.0e5	$9.8e4
GSE16032	343	10	70.26%	0.03 years	0.02 years
				$9.3e4	$6.5e4
GSE16538	1,624	12	14.83%	0.19 years	0.03 years
				$9.3e4	$6.6e4
GSE11524	2,295	30	10.50%	0.68 years	0.07 years
				$7.1e5	$7.5e4

For each data set, differential expression was calculated with respect to the primary biological variable. For each experiment, the number of genes differentially expressed at 5% is reported. In each case, 241 genes are required for statistical validation, for each study we present the fraction of the DE genes required for statistical validation. Tthe cost in dollars and graduate student years of manually confirming the whole list of DE genes or only the DE genes needed for statistical validation is also reported.

ISSN: 1471-2105