GFS: fuzzy preprocessing for effective gene expression analysis

BMC Bioinformatics

Table 4 Leukemia – Significance comparison of size of subnetworks induced by high-variance genes in preprocessed output; p ₁ = p-value using first three PCs, p ₂ = p-value using PC2, PC3 only

	Raw			Scaled			Z-score			Quantile			GFS
Size	freq	p ₁	p ₂	freq	p ₁	p ₂	freq	p ₁	p ₂	freq	p ₁	p ₂	freq	p ₁	p ₂
2	44	0.994	0.993	57	0.920	0.924	47	0.987	0.986	59	0.893	0.894	47	0.570	0.582
3	34	0.557	0.584	23	0.954	0.945	27	0.842	0.844	26	0.885	0.883	34	0.073	0.075
4	14	0.664	0.679	14	0.664	0.679	15	0.588	0.589	17	0.454	0.471	15	0.046	0.044
5	7	0.597	0.579	6	0.700	0.686	8	0.474	0.465	7	0.597	0.579	4	0.244	0.253
6	5	0.279	0.318	2	0.762	0.779	1	0.904	0.925	4	0.423	0.462	6	0.011	0.013
7	1	0.688	0.696	1	0.688	0.696	4	0.130	0.149	1	0.688	0.696	2	0.159	0.166
8	4	0.048	0.039	3	0.119	0.104	-	-	-	1	0.487	0.500	1	0.259	0.220
9	1	0.384	0.369	2	0.153	0.159	3	0.051	0.047	4	0.014	0.011	3	0.021	0.017
10	1	0.285	0.252	2	0.107	0.098	1	0.285	0.252	1	0.285	0.252	2	0.032	0.031
11	1	0.201	0.224	-	-	-	-	-	-	1	0.201	0.224	2	0.020	0.017
12	-	-	-	-	-	-	-	-	-	-	-	-	1	0.030	0.028
15	-	-	-	-	-	-	-	-	-	-	-	-	1	0.006	0.001

ISSN: 1471-2105