Skip to main content

Table 1 Validation dataset accuracy for MetaVelvet-DL and MetaVelvet-SL

From: MetaVelvet-DL: a MetaVelvet deep learning extension for de novo metagenome assembly

  MetaVelvet-DL MetaVelvet-SL
Accuracy 78.3% 57.5%
Class 1 Sensitivity 75.5% 64.3%
Class 1 Specificity 90.0% 80.7%
Class 2 Sensitivity 85.4% 66.8%
Class 2 Specificity 86.2% 71.0%
Class 3 Sensitivity 73.9% 41.4%
Class 3 Specificity 91.3% 84.7%