Skip to main content
. 2017 Oct 23;34(7):1092–1098. doi: 10.1093/bioinformatics/btx662

Table 3.

Prediction performance of PaRSnIP compared with six protein solubility prediction tools

PaRSnIP PROSO II CCSOL SOLpro PROSO RPSP SCM
Accuracy (%) 74.11 64.35 54.20 59.95 57.85 51.45 59.67
MCC 0.48 0.31 0.08 0.20 0.16 0.03 0.21
Sensitivity (soluble) 0.73 0.46 0.51 0.51 0.54 0.44 0.42
Sensitivity (insoluble) 0.75 0.82 0.57 0.69 0.62 0.59 0.77
Selectivity (soluble) 0.75 0.72 0.54 0.62 0.58 0.52 0.65
Selectivity (insoluble) 0.74 0.60 0.54 0.58 0.57 0.51 0.57
Gain (soluble) 1.50 1.45 1.09 1.24 1.17 1.03 1.30
Gain (insoluble) 1.47 1.21 1.08 1.17 1.15 1.02 1.14

Note: Best performing method in bold. Performance values adopted from Chang et al. (2014).