Table 2.
Performance of eight low-frequency variant callers on simulated datasets at various VAFs of 5%, 2.5%, 1%, 0.5%, 0.25%, 0.1%, 0.05%, and 0.025% with a sequencing depth of 20000X.
| Variant caller | VAF = 5% | VAF = 2.5% | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| No. of calls a | True positive | False positive | Sensitivity (%) | Precision (%) | No. of calls | True positive | False positive | Sensitivity (%) | Precision (%) | |
| DeepSNVMiner | 44 | 44 | 0 | 88.00 | 100.00 | 44 | 44 | 0 | 88.00 | 100.00 |
| MAGERI | 41 | 41 | 0 | 82.00 | 100.00 | 41 | 41 | 0 | 82.00 | 100.00 |
| smCounter2 | 50 | 49 | 1 | 98.00 | 98.00 | 50 | 49 | 1 | 98.00 | 98.00 |
| UMI-VarCal | 48 | 48 | 0 | 96.00 | 100.00 | 48 | 48 | 0 | 96.00 | 100.00 |
| SiNVICT | 56 | 49 | 7 | 98.00 | 87.50 | 56 | 49 | 7 | 98.00 | 87.50 |
| outLyzer | 52 | 50 | 2 | 100.00 | 96.15 | 52 | 50 | 2 | 100.00 | 96.15 |
| Pisces | 63 | 49 | 14 | 98.00 | 77.78 | 63 | 49 | 14 | 98.00 | 77.78 |
| LoFreq | 91 | 48 | 43 | 96.00 | 52.75 | 91 | 48 | 43 | 96.00 | 52.75 |
| Variant caller | VAF = 1% | VAF = 0.5% | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| No. of calls | True positive | False positive | Sensitivity (%) | Precision (%) | No. of calls | True positive | False positive | Sensitivity (%) | Precision (%) | |
| DeepSNVMiner | 44 | 44 | 0 | 88.00 | 100.00 | 44 | 44 | 0 | 88.00 | 100.00 |
| MAGERI | 41 | 41 | 0 | 82.00 | 100.00 | 41 | 41 | 0 | 82.00 | 100.00 |
| smCounter2 | 0 | 0 | 0 | 0.00 | 0.00 | 0 | 0 | 0 | 0.00 | 0.00 |
| UMI-VarCal | 48 | 48 | 0 | 96.00 | 100.00 | 48 | 48 | 0 | 96.00 | 100.00 |
| SiNVICT | 8 | 1 | 7 | 2.00 | 12.50 | 7 | 0 | 7 | 0.00 | 0.00 |
| outLyzer | 3 | 1 | 2 | 2.00 | 33.33 | 2 | 0 | 2 | 0.00 | 0.00 |
| Pisces | 21 | 7 | 14 | 14.00 | 33.33 | 14 | 0 | 14 | 0.00 | 0.00 |
| LoFreq | 88 | 20 | 68 | 40.00 | 22.73 | 82 | 8 | 74 | 16.00 | 9.76 |
| Variant caller | VAF = 0.25% | VAF = 0.1% | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| No. of calls | True positive | False positive | Sensitivity (%) | Precision (%) | No. of calls | True positive | False positive | Sensitivity (%) | Precision (%) | |
| DeepSNVMiner | 44 | 44 | 0 | 88.00 | 100.00 | 43 | 43 | 0 | 86.00 | 100.00 |
| MAGERI | 40 | 40 | 0 | 80.00 | 100.00 | 39 | 39 | 0 | 78.00 | 100.00 |
| smCounter2 | 0 | 0 | 0 | 0.00 | 0.00 | 0 | 0 | 0 | 0.00 | 0.00 |
| UMI-VarCal | 47 | 47 | 0 | 94.00 | 100.00 | 46 | 46 | 0 | 92.00 | 100.00 |
| SiNVICT | 8 | 0 | 8 | 0.00 | 0.00 | 7 | 0 | 7 | 0.00 | 0.00 |
| outLyzer | 2 | 0 | 2 | 0.00 | 0.00 | 2 | 0 | 2 | 0.00 | 0.00 |
| Pisces | 14 | 0 | 14 | 0.00 | 0.00 | 14 | 0 | 14 | 0.00 | 0.00 |
| LoFreq | 72 | 3 | 69 | 6.00 | 4.17 | 83 | 0 | 83 | 0.00 | 0.00 |
| Variant caller | VAF = 0.05% | VAF = 0.025% | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| No. of calls | True positive | False positive | Sensitivity (%) | Precision (%) | No. of calls | True positive | False positive | Sensitivity (%) | Precision (%) | |
| DeepSNVMiner | 44 | 44 | 0 | 88.00 | 100.00 | 0 | 0 | 0 | 0.00 | 0.00 |
| MAGERI | 39 | 39 | 0 | 78.00 | 100.00 | 0 | 0 | 0 | 0.00 | 0.00 |
| smCounter2 | 0 | 0 | 0 | 0.00 | 0.00 | 0 | 0 | 0 | 0.00 | 0.00 |
| UMI-VarCal | 42 | 42 | 0 | 84.00 | 100.00 | 0 | 0 | 0 | 0.00 | 0.00 |
| SiNVICT | 7 | 0 | 7 | 0.00 | 0.00 | 7 | 0 | 7 | 0.00 | 0.00 |
| outLyzer | 2 | 0 | 2 | 0.00 | 0.00 | 2 | 0 | 2 | 0.00 | 0.00 |
| Pisces | 14 | 0 | 14 | 0.00 | 0.00 | 14 | 0 | 14 | 0.00 | 0.00 |
| LoFreq | 76 | 0 | 76 | 0.00 | 0.00 | 80 | 0 | 80 | 0.00 | 0.00 |
anumber of calls.
Significant values are highlighted in bold.