Table 4.
SV genotyping method | Aligner | Sequencing data | Max-Min | ||
---|---|---|---|---|---|
CLR | ONT | CCS | |||
cuteSV | minimap2 | 0.89 | 0.93 | 0.93 | 0.06 |
NGMLR | 0.87 | 0.91 | 0.91 | ||
LRcaller | minimap2 | 0.86 | 0.85 | 0.88 | 0.03 |
NGMLR | 0.87 | 0.87 | 0.87 | ||
Sniffles | minimap2 | 0.68 | 0.81 | 0.83 | 0.18 |
NGMLR | 0.65 | 0.73 | 0.76 | ||
SVJedi | minimap2 | 0.86 | 0.81 | NA | 0.05 |
VaPoR | minimap2 | 0.67 | 0.71 | 0.71 | 0.04 |
NGMLR | 0.67 | 0.71 | 0.71 |
Impact of aligner (minimap2 and NGMLR) and sequencing data (PacBio CLR, PacBio CCS, and ONT) on the F1 score of each genotyping method based on the HG002 Tier 1 dataset. Performance was evaluated on 30× HG002 sequencing data. SVJedi does not support the output from NGMLR. “NA” indicates the data is not available. The bold black number is the highest F1 score for each SV genotyping method. The “Max-Min” column represents the maximum F1 score minus the minimum F1 score for each SV genotyping method under different combinations of aligners and sequencing data