Table 2. Dependence of the prediction performance on the size of positive training sample set.

Percentage (%) of positive training examples	AUC of MNMC.slif
100	0.7298
90	0.7332
80	0.7289
70	0.7328
60	0.7261
50	0.7211
40	0.7187
30	0.7051
20	0.7125
10	0.7019

Different portions (10% to 100%) of the 9994 SL examples in our data set were used in a 10-fold cross-validation procedure to test the efficacy of the resultant predictor.