Table 3.
Comparison of two representation methods for predicting human proteins interacting with HCV proteins
| Test set | Our representation | Shen's representation | ||||
|---|---|---|---|---|---|---|
| SN (%) | SP (%) | AC (%) | SN (%) | SP (%) | AC (%) | |
| 1 | 75.9 | 83.6 | 79.7 | 73.8 | 82.1 | 77.9 |
| 2 | 80.5 | 89.7 | 85.1 | 73.8 | 82.1 | 77.9 |
| 3 | 76.9 | 83.1 | 80 | 74.4 | 76.9 | 75.6 |
| Average | 77.8 | 85.4 | 81.6 | 74.0 | 80.4 | 77.1 |
Comparison of our representation method with the method by Shen et al. [5] in terms of sensitivity (SN), specificity (SP) and accuracy (AC) with the HCV data. The two representation methods are different in their classification of 20 amino acids and definition of the relative frequency of an amino acid triplet (see Methods section for details).