Skip to main content

Table 4 Performance with different parameter settings

From: De novo Nanopore read quality improvement using deep learning

  Default (Length, Depth) = (36, 36) (w,k) = (7,17)
Mean Sq. Error 0.00300 0.00329 0.00305
Pearson 0.827 0.821 0.830
Spearman 0.805 0.786 0.810
Sensitivity 0.950 0.934 0.914
Specificity 0.681 0.693 0.780
  1. Performance with different parameter settings. w and k refer to the minimizer parameters, while Length and Depth refer to the length and depth of each pileup image segment, which correspond to the number of minimizers in that read segment and the number of matching reads used. The default settings are (w,k) =(5,15) and (Length, Depth) = (48, 24). The columns show the performance when varying one of these settings and with cutoff 0.8