Distributional fold change test – a statistical approach for detecting differential expression in microarray experiments

Table 2 AUC performance of DFC test, t-test, and shrinkage t-test

GEO data set	N _s	AUC for MAS5 pre-processed data			AUC for RMA pre-processed data
		t-test	ShrinkT^a	DFC	t-test	ShrinkT^a	DFC
GSE8441	22	0.92912	0.94404	0.96996	0.91206	0.92842	0.96812
GSE9499	22	0.96425	0.98255	0.98529	0.94735	0.97241	0.9718
GSE2639	14	0.99782	0.99838	0.9987	0.99851	0.99784	0.99896
GSE2638	7	0.79197	0.83621	0.86199	0.75527	0.82421	0.83175
GSE3860	18	0.98986	0.99581	0.99742	0.98647	0.99246	0.99568
GSE6344	20	0.97165	0.98078	0.98854	0.97586	0.98216	0.9889
GSE7765	6	0.96323	0.97846	0.98564	0.96267	0.98146	0.98939
GSE6740_1	20	0.99491	0.99676	0.99701	0.9972	0.99803	0.99803
GSE6740_2	20	0.99115	0.99313	0.99283	0.97599	0.98248	0.98487
GSE6011	37	0.86072	0.8674	0.90942	0.97544	0.98126	0.97892
GSE2531	7	0.91614	0.94288	0.9379	0.93889	0.94368	0.94107
Average^b		0.9718	0.9812	0.9857	0.9745	0.9815	0.9861

AUC performance of DFC test, t-test, and shrinkage t-test on MAS5 and RMA pre-processed data from data sets described in Table1. N_s is the number of samples in the set. ^aShrinkT -test values were calculated with CAT-test[14], option ‘diagonal’. ^bAverage was calculated for logit transformed AUC values, LTA = 0.5⋅ln(AUC/(1-AUC)) and then transformed back to AUC scale.

ISSN: 1748-7188