Table 1. Breast cancer patient cohorts used for biomarker discovery and validation.

Throughout this document we will use the definition of HR negative as ER negative and PR negative. While HR positive is defined as not HR negative, i.e. ER positive and PR positive, ER positive and PR negative, ER negative and PR positive.

	Discovery		Validation
Data	Hatzis et al.	Hatzis et al.	Horak et al.	MAQC consortium
Name alias	-	hatzis182	horak121	maqc264
Source	E-GEOD-25055	E-GEOD-25065	E-GEOD-41998	E-GEOD-20194
Platform	HG-U133A	HG-U133A	HG-U133A_2	HG-U133A
N(1)	306	182	121	264
HR status
positive	184	121	55	-
negative	117	60	66	-
ER status
positive	172	113	45	161
negative	129	68	76	103
PR status
positive	140	94	46	-
negative	160	87	75	-
HER2 status(2)
positive	3	0	9	56
negative	288	182	112	208
Response
pCR	57	42	34	55
RD	249	140	87	209
Response Rate	19%	23%	28%	20%
Chemotherapy regime(3)
neoadjuvant	306	165	121	264
partial adjuvant	0	18	0	0
adjuvant	0	15	0	0
Taxane(4)
Paclitaxel	287	92	121	264
Docetaxel	18	90	0	0
Combination
FAC (5)	227	103	0	182
AC (6)	83	0	121	0
FEC (7)	0	125	0	78
X (8)	0	94	0	0
Trastuzumab	0	0	0	8
Other	0	0	0	5

⁽¹⁾ Patients with reported response status were considered

⁽²⁾ Samples of the Hatzis et al. study are assumed HER2- where no explicit meta information has been found

⁽³⁾ Chemotherapy regime was not reported for individual patients by Hatzis et al.

⁽⁴⁾ Taxane was not specified for one patient in Discovery

⁽⁵⁾ Fluorouracil (F), doxorubicin (A) and cyclophosphamide (C).

⁽⁶⁾ Doxorubicin (A) and cyclophosphamide (C).

⁽⁷⁾ Fluorouracil (F), epirubicin (E) and cyclophosphamide(C).

⁽⁸⁾ Capecitabine.