From: iEnhancer-DCLA: using the original sequence to identify enhancers and their strength based on a deep learning framework
Stages
Model
Acc(%)
Sn(%)
Sp(%)
MCC
First layer
CNN
70.50
92.00
49.00
0.4541
BiLSTM
73.00
89.00
57.00
0.4855
CNN-BiLSTM
75.50
59.00
CNN-BiLSTM-Attention
78.00
87.00
69.00
0.5693