Table 10 Comparison of different methods on three datasets.
From: Hybrid framework for membrane protein type prediction based on the PSSM
Method | Dataset1 | Dataset2 | Dataset3 | ||||
|---|---|---|---|---|---|---|---|
Validation set | Test set | Validation set | Test set | Validation set | Test set | ||
Comparison method | PsePSSM_Ensemble | – | 91.6 | – | 78.3 | – | 85.3 |
Physicochemical-Ensemble | – | 91.0 | – | – | – | – | |
Integrate-multi_base Pseacc | – | 92.6 | – | 88.2 | – | 90.3 | |
PsePSSM-LLDA | – | 88.7 | – | – | – | – | |
PsePSSM-DC | – | 90.2 | – | – | – | – | |
PsePSSM-PCA | – | 80.7 | – | – | – | – | |
Fea-fusion | – | 93.5 | – | – | – | – | |
Ave-WT | – | 92.7 | – | 88.1 | – | 89.6 | |
MKSVM-HSIC | – | 94.4 | – | 89.8 | – | 90.7 | |
SIM | 85.5 | 91.6 | 83.2 | 89.1 | – | – | |
Capnet | 90.5 | 94.0 | 81.0 | 89.8 | 84.1 | 90.0 | |
RNN | – | 94.1 | – | 89.4 | – | – | |
Our method | OCNN | 92.9 | 94.6 | 83.4 | 88.4 | 88.6 | 92.1 |
ICNN | 94.8 | 95.7 | 83.9 | 90.5 | 90.8 | 92.8 | |
Hybrid | – | 95.8 | – | 92.0 | – | 93.4 | |
