Table 3.

Evaluation of cell cycle gene expression signature as breast cancer prognosis markers by supervised methods.

		Dataset
Gene signature model	Number of genes used in the classification model	Training and testing: Wang dataset [8]	Independent validation: Miller dataset [22]	Independent validation: Pawitan dataset [7]
The 70-gene signature	51	6.1E-05	0.057	0.051
Breast cancer	232	2.6E-09	0.0012	0.0019
Cell Cycle	108	1.4E-06	0.0050	0.0046
Random	232	1.8E-13	0.14	0.52

The numbers represent the log-rank test P values in Kaplan-Meier analysis in the good and poor prognosis groups predicted by the Amsterdam 70-gene signature, the breast cancer gene set, the cell cycle gene set classifier, and the randomly selected gene set respectively.