Table 2 Comparison of base-calling performances between DNBSRN and other networks.
From: Deep learning enables the use of ultra-high-density array in DNBSEQ
Data | Metrics | EDSR | RDN | RCAN | IMDN | RFDN | RLFN | DNBSRN | |
|---|---|---|---|---|---|---|---|---|---|
SE50 | Dataset1 | Q30(%) | 77.91 | 77.87 | 78.16 | 78.14 | 77.48 | 77.56 | 78.25 |
MR(%) | 82.36 | 81.93 | 82.57 | 82.79 | 82.11 | 82.35 | 83.07 | ||
ESR(%) | 79.76 | 79.90 | 80.45 | 80.37 | 80.17 | 80.33 | 80.54 | ||
Dataset2 | Q30(%) | 76.22 | 76.24 | 76.32 | 76.18 | 75.68 | 75.78 | 76.45 | |
MR(%) | 78.31 | 78.35 | 78.57 | 78.87 | 78.58 | 78.85 | 79.38 | ||
ESR(%) | 77.21 | 77.44 | 77.69 | 77.62 | 77.42 | 77.55 | 78.01 | ||
Dataset3 | Q30(%) | 76.91 | 76.99 | 76.97 | 76.74 | 76.08 | 75.97 | 76.94 | |
MR(%) | 79.31 | 79.57 | 79.79 | 79.92 | 79.30 | 79.46 | 80.18 | ||
ESR(%) | 77.74 | 78.17 | 78.36 | 78.12 | 77.71 | 77.63 | 78.30 | ||
Dataset4 | Q30(%) | 81.18 | 81.27 | 81.22 | 81.14 | 81.11 | 80.82 | 81.49 | |
MR(%) | 87.38 | 87.25 | 87.34 | 87.15 | 87.18 | 86.92 | 87.55 | ||
ESR(%) | 82.48 | 82.56 | 82.62 | 82.38 | 82.51 | 82.21 | 82.73 | ||
PE100 | Dataset5 | Q30(%) | 74.73 | 73.96 | 75.19 | 75.07 | 74.15 | 74.56 | 75.54 |
MR(%) | 83.35 | 82.42 | 83.59 | 83.95 | 82.26 | 83.07 | 84.76 | ||
ESR(%) | 76.70 | 75.91 | 77.68 | 77.81 | 77.67 | 78.07 | 78.30 | ||
Dataset6 | Q30(%) | 74.30 | 73.52 | 74.78 | 74.78 | 73.83 | 74.25 | 75.24 | |
MR(%) | 82.93 | 81.92 | 83.25 | 83.64 | 81.98 | 82.73 | 84.49 | ||
ESR(%) | 76.31 | 75.50 | 77.34 | 77.61 | 77.44 | 77.85 | 78.08 | ||
Dataset7 | Q30(%) | 74.77 | 74.01 | 75.20 | 75.07 | 74.17 | 74.56 | 75.55 | |
MR(%) | 83.27 | 82.33 | 83.52 | 83.85 | 82.27 | 83.02 | 84.64 | ||
ESR(%) | 76.83 | 76.01 | 77.72 | 77.89 | 77.76 | 78.16 | 78.37 | ||
Dataset8 | Q30(%) | 74.61 | 73.79 | 74.91 | 74.94 | 74.01 | 74.41 | 75.41 | |
MR(%) | 83.10 | 82.12 | 83.21 | 83.71 | 82.07 | 82.82 | 84.53 | ||
ESR(%) | 76.68 | 75.77 | 77.47 | 77.74 | 77.59 | 77.99 | 78.24 |
