Skip to main content
. 2019 Nov 15;20(22):5743. doi: 10.3390/ijms20225743

Table 5.

Performance comparisons between Meta-iAVP and the three existing methods as assessed by the five-repeated five-fold cross-validation and independent validation tests.

Dataset Method a Ac (%) Sn (%) Sp (%) MCC
T544p+407n AVPpred 85.00 82.20 88.20 0.70
Chang et al.’s method 85.10 86.60 83.00 0.70
AntiVPP 1.0 - - - -
Meta-iAVP 88.20 89.20 86.90 0.76
T544p+544n AVPpred 90.00 89.70 90.30 0.80
Chang et al.’s method 91.50 89.00 94.10 0.83
AntiVPP 1.0 - - - -
Meta-iAVP 93.20 89.00 97.40 0.87
V60p+45n AVPpred 85.70 88.30 82.20 0.71
Chang et al.’s method 89.50 91.70 86.70 0.79
AntiVPP 1.0 - - - -
Meta-iAVP 95.20 96.70 93.20 0.90
V60p+60n AVPpred 92.50 93.30 91.70 0.85
Chang et al.’s method 93.30 91.70 95.00 0.87
AntiVPP 1.0 93.00 87.00 97.00 0.87
Meta-iAVP 94.90 91.70 98.30 0.90

a Results were reported from the works of AVPpred, Chang et al.’s method, and AntiVPP 1.0. The highest values for each performance measure are shown in bold.