Research Article

Detecting Protein-Protein Interactions with a Novel Matrix-Based Protein Sequence Representation and Support Vector Machines

Table 2

Comparing the prediction performance by the proposed method and some state-of-the-art works on the yeast dataset. Here, N/A means not available.

ModelTest setSN (%)PPV (%)ACC (%)MCC (%)

Proposed methodSVM85.74 ± 0.9493.84 ± 0.9890.06 ± 0.6482.03 ± 1.03

Guos’ workACC89.93 ± 3.6888.87 ± 6.1689.33 ± 2.67N/A
AC87.30 ± 4.6887.82 ± 4.3387.36 ± 1.38N/A

Zhous’ workSVM + LD87.37 ± 0.2289.50 ± 0.6088.56 ± 0.3377.15 ± 0.68

Yangs’ workCod175.81 ± 1.2074.75 ± 1.2375.08 ± 1.13N/A
Cod276.77 ± 0.6982.17 ± 1.3580.04 ± 1.06N/A
Cod378.14 ± 0.9081.86 ± 0.9980.41 ± 0.47N/A
Cod481.03 ± 1.7490.24 ± 1.3486.15 ± 1.17N/A