Skip to main content

Table 5 The performance of model reduction

From: SigUNet: signal peptide recognition based on semantic segmentation

Method Eukaryotes Gram-positive Gram-negative
MCC (%) FPRTM (%) MCC (%) FPRTM (%) MCC (%) FPRTM (%)
The SignalP dataset
 SigUNet 90.2 4.0 76.1 5.1 80.6 1.5
 SigUNet-light 89.4 4.3 77.7 5.1 82.9 1.9
The SPDS17 dataset
 SigUNet 89.6 1.2 40.9 2.1 97.6 1.1
 SigUNet-light 84.8 3.7 51.7 1.6 92.8 1.1
  1. Performances that are improved after model reduction are highlighted in bold.