Table 4 |
|||||||
|
The performance of SVM model on realistic dataset using SRT, DPT and TPT (See Additional file 1 Table S7-12 for detail) |
|||||||
|
Types of patterns |
Method |
Sensitivity (Recall) |
Specificity |
Accuracy |
Precision |
MCC |
F1 Score |
|
|
|||||||
|
SRT |
Single Sequence |
24.00 |
99.36 |
95.65 |
0.66 |
0.38 |
0.36 |
|
PSSM profile |
41.65 |
99.55 |
96.57 |
0.83 |
0.57 |
0.56 |
|
|
DPT |
Single Sequence |
51.22 |
93.52 |
89.65 |
0.44 |
0.42 |
0.48 |
|
PSSM profile |
62.31 |
98.63 |
95.33 |
0.82 |
0.69 |
0.71 |
|
|
TPT |
Single Sequence |
50.39 |
94.60 |
90.61 |
0.48 |
0.44 |
0.49 |
|
PSSM profile |
66.80 |
99.11 |
96.30 |
0.88 |
0.75 |
0.76 |
|
|
|
|||||||
|
Where SRT: Single Residue based Technique DPT: Dipeptide based Technique TPT: Tripeptide based Technique |
|||||||
|
Chauhan et al. BMC Bioinformatics 2010 11:301 doi:10.1186/1471-2105-11-301 |
|||||||
