Modified Logistic Regression Models Using Gene Coexpression and Clinical Features to Predict Prostate Cancer Progression

Table 2

Comparison of the performance of our logistic regression models with that of the nine models evaluated by Sboner et al. [5], using the same number of genes.

Model number	Patient’s age	Gleason score	Tumor percentage	Fusion ERG	Number of genes	AUC in ref. [5]	AUC of our model
2.1				X	18	0.672	0.769
2.2	X			X	9	0.708	0.732
2.3					18	0.713	0.736
2.4		X		X	21	0.726	0.793
2.5	X				11	0.730	0.712
2.6	X	X	X		3	0.738	0.806
2.7	X	X		X	12	0.745	0.804
2.8		X			16	0.749	0.813
2.9	X	X			12	0.750	0.788