Skip to main content

Table 1 AUC, Success rate and FPR for varying flank size with RBF and sigmoid kernels (C value used is 1000)

From: Predicting MoRFs in protein sequences using HMM profiles

  AUC   Success rate   FPR @ 0.222 TPR  
RBF kernel Sigmoid kernel RBF kernel Sigmoid kernel RBF kernel Sigmoid kernel
Gamma 0.0038 5 0.0038 5 0.0038 5 0.0038 5 0.0038 5 0.0038 5
Flank size
1 0.658 0.587 0.570 0.648 0.680 0.660 0.658 0.701 0.057 0.160 0.090 0.070
2 0.659 0.597 0.590 0.653 0.651 0.737 0.653 0.680 0.053 0.190 0.088 0.065
3 0.650 0.600 0.580 0.650 0.640 0.770 0.660 0.660 0.047 0.180 0.080 0.065
4 0.660 0.606 0.580 0.340 0.640 0.770 0.589 0.370 0.053 0.180 0.090 0.380
5 0.660 0.600 0.587 0.650 0.669 0.720 0.618 0.600 0.050 0.180 0.098 0.060
6 0.659 0.600 0.589 0.648 0.649 0.749 0.618 0.572 0.053 0.180 0.098 0.066
7 0.664 0.601 0.588 0.340 0.644 0.756 0.642 0.460 0.051 0.170 0.090 0.360
8 0.652 0.602 0.595 0.350 0.653 0.740 0.600 0.470 0.059 0.170 0.095 0.360
9 0.646 0.584 0.582 0.618 0.653 0.699 0.584 0.390 0.061 0.180 0.010 0.073
10 0.644 0.587 0.640 0.590 0.656 0.699 0.432 0.590 0.065 0.175 0.077 0.100
11 0.645 0.605 0.640 0.639 0.668 0.749 0.604 0.390 0.066 0.160 0.105 0.080
12 0.640 0.600 0.600 0.630 0.670 0.810 0.630 0..36 0.070 0.160 0.090 0.080
  1. Bold numbers indicate the best performance metrics for different kernels, gamma values and Flank sizes