Skip to main content

Table 7 Comparison to other protein disorder predictors over VSL2 training dataset via 10-fold cross-validation. Predictions by VSL2B, VSL2P and VSL2 were made via the 10-fold cross-validation procedure, while other predictors were applied to the 1,327 sequences directly. Default thresholds were used for all predictors, e.g. 0.5 for VL-XT, VL3-E, VSL1, VSL2B, VSL2P and VSL2. Also shown are the areas under ROC curves (AUC) in Figure 5.

From: Length-dependent prediction of protein intrinsic disorder

  SN SP ACC SN S SN L AUC
VL-XT 58.6 ± 1.3 78.6 ± 0.4 68.6 ± 0.7 56.3 ± 1.6 63.5 ± 2.1 75.7 ± 0.8
VL3-E 38.7 ± 1.6 92.7 ± 0.5 65.7 ± 0.8 23.5 ± 1.6 82.5 ± 2.2 76.0 ± 0.8
DisEMBL 31.4 ± 1.4 97.8 ± 0.1 64.6 ± 0.7 30.5 ± 1.6 32.8 ± 2.2 84.5 ± 0.6
RONN 45.8 ± 1.4 87.0 ± 0.4 66.4 ± 0.7 39.6 ± 1.7 63.0 ± 2.3 72.2 ± 0.9
DISOPRED2 56.9 ± 1.3 94.1 ± 0.2 75.5 ± 0.7 56.5 ± 1.6 55.1 ± 2.4 85.7 ± 0.6
VSL1 79.0 ± 1.1 85.3 ± 0.4 82.2 ± 0.6 78.0 ± 1.3 78.1 ± 2.0 89.9 ± 0.6
VSL2B 77.3 ± 1.1 79.9 ± 0.4 78.6 ± 0.6 75.8 ± 1.3 78.2 ± 1.9 86.0 ± 0.6
VSL2P 81.0 ± 1.0 80.4 ± 0.5 80.7 ± 0.6 79.8 ± 1.2 80.9 ± 1.9 88.0 ± 0.6
VSL2 82.3 ± 1.1 81.0 ± 0.5 81.6 ± 0.5 81.3 ± 1.2 82.3 ± 1.8 89.2 ± 0.5
(a) per-chain
  SN SP ACC SN S SN L AUC
VL-XT 58.9 ± 1.7 79.2 ± 0.3 69.0 ± 0.9 51.4 ± 1.7 61.6 ± 2.2 75.7 ± 1.2
VL3-E 70.4 ± 3.1 94.5 ± 0.4 82.5 ± 1.6 27.2 ± 1.9 85.7 ± 2.7 90.9 ± 1.0
DisEMBL 32.5 ± 2.0 98.2 ± 0.1 65.4 ± 1.0 28.7 ± 1.5 33.9 ± 2.7 80.0 ± 1.1
RONN 61.1 ± 2.9 87.6 ± 0.3 74.4 ± 1.5 42.7 ± 1.9 67.6 ± 3.3 81.5 ± 1.5
DISOPRED2 60.2 ± 3.7 95.1 ± 0.2 77.6 ± 1.8 50.1 ± 1.7 63.7 ± 4.6 87.7 ± 1.2
VSL1 78.1 ± 2.3 86.7 ± 0.3 82.4 ± 1.2 71.4 ± 1.5 80.4 ± 2.9 90.3 ± 1.0
VSL2B 77.0 ± 2.2 81.5 ± 0.3 79.3 ± 1.1 67.6 ± 1.6 80.4 ± 2.7 87.1 ± 1.2
VSL2P 81.7 ± 2.2 82.2 ± 0.4 81.9 ± 1.1 75.6 ± 1.5 83.8 ± 2.7 89.8 ± 1.2
VSL2 82.9 ± 2.1 81.6 ± 0.4 82.3 ± 1.1 77.6 ± 1.4 84.7 ± 2.7 90.5 ± 1.1
(b) per-residue