Table 8.
Evaluation of the correction of a real human dataset with ELECTOR
| Uncorrected | Corrected with MECAT | |
|---|---|---|
| Processed bases | 5 605 157 590 | 5 451 767 836 |
| Recall (%) | – | 92.70 |
| Precision (%) | – | 91.50 |
| Error rate | 0.1974 | 0.0861 |
| Trimmed/split | – | 570 635 |
| Mean missing size | – | 362.0 |
| Extended | – | 275 |
| Mean extension size | – | 62.4 |
| Low quality reads | – | 4279 |
| Small reads | – | 356 |
| Insertions | 247 953 086 | 10 144 736 |
| Deletions | 746 165 024 | 473 239 036 |
| Substitutions | 162 822 923 | 7 521 389 |
| Homopolymer ratio | – | 0.7570 |
| Runtime | – | 18 h 27 min |
The reads were corrected with MECAT, using default parameters, before the evaluation. ELECTOR evaluated a total of 650 771 reads. Small reads are corrected reads whose length is <10% of the original read. Low quality corrected reads are reads for which an insufficient number of seeds was found during the segmentation process. Homopolymer ratio is the ratio of homopolymer sizes in corrected versus reference. We reported the wallclock time of the run, using 20 threads.