Table 1. Description of the 12 data sets used in our study.

The data sets are obtained from GEO. Each entry has the GEO code the name of the data, the number of samples and the classes of the data.

GEO Accession	Title	Sample count	Classes
GDS1962	Glioma-derived stem cell factor effect on angiogenesis in the brain	180 pos=157 neg=23	non-tumor=23 (neg) astrocytomas=26 (pos) glioblastomas=131 (pos)
GDS2519	Early-stage Parkinson's disease: whole blood	105 pos=50 neg=55	healthy control=22 (neg) neurodegenerative disease control=33 (neg) Parkinson disease=50 (pos)
GDS3268	Colon epithelial biopsies of ulcerative colitis patients	202 pos=73 neg=129	normal=73 ulcerative colitis=129
GDS2547	Metastatic prostate cancer (HG-U95C)	164 pos=75 neg=89	normal=75 tumor=89
GDS5499	Pulmonary hypertensions: PBMCs	140 pos=99 neg=41	control=41 (neg) idiopathic pulmonary arterial hypertension=30 (pos) scleroderma-associated pulm. arterial hypert=42 (pos) systemic sclerosis (SSc) without pulm. hypert=19 (pos) SSc, interstitial lung disease & pulm. hypert=8 (pos)
GDS3646	Celiac disease: primary leukocytes	132 pos=110 neg=22	healthy control=22 celiac disease=110
GDS3874	Diabetic children: peripheral blood mononuclear cells (U133A)	117 pos=93 neg=24	healthy=24 type 1,2 diabetes=93
GDS3837	Non-small cell lung carcinoma in female nonsmokers	120 pos=60 neg=60	lung cancer=60 control=60
GDS5037	Severe asthma: bronchial epithelial cell	108 pos=88 neg=20	mild asthma=50 control=20 severe asthma=38
GDS4516_ 4718	Colorectal cancer: laser microdissected tumor tissues Colorectal cancer: homogenized tumor tissues	148 pos=104 neg=44	laser microdissected tumor tissues=104 homogenized tumor tissues=44
GDS3900	Fear conditioning effect on the hybrid mouse diversity panel: hippocampus and striatum	198 pos=100 neg=98	hippocampus=100 (pos) striatum=98 (neg)
GDS3929	Tobacco smoke effect on maternal and fetal cells	183 pos=128 neg=55	non-smoker=128 (pos) smoker=55 (neg)