Skip to main content
. 2021 Jan 4;12:60. doi: 10.1038/s41467-020-20236-7

Table 1.

Performance comparison of nanopore read error correction.

Data sets Pipeline Size (g)/time (h)/speed (g/h) Error rate (%) ≤5%(%) N50 N75 Read number with HERS
E. coli raw reads 1.38/–/– 17.8 0.01 41,074 35,484 121
Canu 0.22/1.63/0.14 7.06 20.45 37,747 32,127 1
NECAT 1.41/0.76/1.86 2.23 (4.27) 99.34 (80.51) 43,140 37,502 1
S. cerevisiae raw reads 5.48/–/– 12 1.61 34,668 28,152 7589
Canu 2.18/30.83/0.071 3.13 87.3 10,554 4567 4820
NECAT 4.57/3.90/1.17 1.53 (3.08) 95.04 (88.09) 31,364 24,480 268
D. melanogaster raw reads 8.30/–/– 16.2 2.3 17,730 13,621 12,438
Canu 4.79/18.10/0.26 8.15 57.57 15,220 10,658 6523
NECAT 7.52/4.20/1.79 4.89 (7.03) 72.03 (64.18) 17,369 13,104 3481
A. thaliana raw reads 3.08/–/– 20.1 1.57 23,386 16,253 14,483
Canu 2.59/12.07/0.22 12.05 8.09 21,472 13,133 8722
NECAT 2.85/1.33/2.14 9.01 (11.35) 45.85 (25.67) 23,600 15,944 7158
C. reinhardtii raw reads 14.84/–/– 15 1.16 54,409 46,812 4231
Canu 4.61/59.40/0.078 5.35 76.05 53,891 45,934 726
NECAT 14.89/11.53/1.29 1.99 (4.40) 95.18 (82.13) 56,427 48,708 278
O. sativa raw reads 63.40/–/– 15.6 0.49 56,325 50,847 24,205
Canu 15.23/43.20/0.35 7.99 44.42 55,010 49,612 4413
NECAT 63.83/18.95/3.37 4.66 (6.45) 74.62 (51.49) 56,573 51,141 3511
S. pennellii raw reads 132.74/–/– 18.49 1.7 24,801 22,226 127,808
Canu 37.53/88.8/0.42 9.69 34.04 21,653 19,364 5511
NECAT 121.07/137.77/0.88 6.45 (9.23) 63.04 (38.77) 23,810 21,480 5445
NA12878 (rel3,4) raw reads 106.52/–/– 18.50 0.67 12,196 7209 286,641
NECAT 101.28/34.65/2.92 5.04 (7.38) 77.60 (34.33) 13,018 7883 53,130
NA12878 (rel6) raw reads 123.80/–/– 12.08 8.91 13,630 7984 315,117
NECAT 98.36/39.35/2.49 6.28 (6.46) 75.45 (77.24) 14,839 9638 64,210

Size is the total number of base pairs in corrected reads. Time is the running time of correction tools, and the speed is the size/time. Error rate denotes the mean error rate of raw reads and corrected reads; ≤5% denotes the percentage of reads with <5% error rate in total corrected read, values in the bracket are results of NECAT after the first correction; N50 and N75 are the length of reads that reached the 50 and 75% of the total length of all reads; read number with HERS denotes the number of reads that with at least one HERS (more than 50% error in the 500 bp window). The reads used in evaluating the last three metrics (N50, N75, and read number with HERS) of NECAT were corrected from the longest 40× of the raw data set that was selected by Canu by default, see Supplementary Note 6 for details.