Table 1.

GEO datasets from the GPL570 platform used for gene expression profiling.

Dataset	Title/Description	Normalization Methods	No. of Samples	Percentage of Cancer
GSE61304	Novel biomarker discovery for stratification and prognosis of breast cancer patients	MAS5 signal intensity	62 (58 breast tumor + 4 normal breast)	94%
GSE42568	Breast cancer gene expression analysis	Log2 GCRMA signal intensity	121 (104 breast tumor + 17 normal breast)	86%
GSE7904	Expression data from human breast tissue	RMA expression value	50 (43 breast tumor + 7 normal breast)	86%
GSE3744	Human breast tumor expression	GCRMA calculated signal intensity, log2 transformed	47 (40 breast tumor + 7 normal breast)	85%
GSE29431	Identifying breast cancer biomarkers	RMA expression values	66 (54 breast tumor + 12 normal breast)	82%
GSE26910	Stromal molecular signatures of breast and prostate cancer	Log2 RMA signal	12 (6 breast tumor + 6 normal breast)	50%
GSE31138	Identifying novel anti-angiogenic targets in human breast cancer	Log2 RMA signal	6 (3 breast tumor + 3 normal breast)	50%
GSE71053	Differential effect of surgical manipulation on gene expression in normal breast tissue and breast tumour tissue	Log2-normalized signal	18 (6 breast tumor + 12 normal breast)	33%
GSE10780	Proliferative genes dominate malignancy risk gene signature in histologically normal breast tissue	RMA expression value	185 (42 breast tumor + 143 normal breast)	23%
GSE30010	Expression data from breast samples of postmenopausal women	RMA expression value	107 (0 breast tumor + 107 normal breast)	0%
GSE111662	Whole breast tissue gene expression in comparison to expression in epithelial and stromal tissues	RMA expression values	27 (0 breast tumor + 27 normal breast)	0%
Total			701 (356 breast tumors + 345 normal breasts)	51%