CoINcIDE: A framework for discovery of patient subtypes across multiple datasets

Table 1 Characteristics of curated breast cancer datasets

Dataset	ID	Batch ID	GEO platform ID	Commercial platform name	Samples (n)	Genes (n)
1	12093		GPL96	Affymetrix Human Genome U133A	136	11,723
2	1379			Arcturus 22 k human oligonucleotide	60	11,723
3	16391			Affymetrix Human Genome U133 Plus 2.0	48	15,199
4	16446			Affymetrix Human Genome U133A	114	16,326
5	17705	JBI		Affymetrix Human Genome U133A	103	10,565
6	17705	MDACC		Affymetrix Human Genome U133A	195	11,026
7	19615			Affymetrix Human Genome U133 Plus 2.0	115	16,652
8	20181			Affymetrix Human Genome U133A	53	10,171
9	20194			Affymetrix Human Genome U133A	261	11,748
10	2034			Affymetrix Human Genome U133A	286	11,020
11	22226			Agilent-012391 Whole Human Genome Oligo G4112A	127	18,841
12	22358			AFFY Human Phase3 v1.0 - C02	121	17,253
13	25055	MDACC_M		Affymetrix Human Genome U133A	221	11,459
14	25065	MDACC		Affymetrix Human Genome U133A	71	11,158
15	25065	USO		Affymetrix Human Genome U133A	54	10,822
16	32646			Affymetrix Human Genome U133 Plus 2.0	115	18,260
17	9893			MLRG Human 21 K V12.0	155	13,154

All breast cancer datasets are from the Gene Expression Omnibus (GEO). ‘Batch dataset’ does not refer to a lab replicate batch, but rather a group of microarrays from a larger study that were run on a different platform, or collected from a different site. These batch labels were inferred from GEO sample file names. GSE = GEO series ID prefix

ISSN: 1756-994X