Table 1.
A summary of eight low-frequency variant callers used in this study.
| Variant caller | Version | Read type | Code | Model | Detection limit (original paper) (%) |
|---|---|---|---|---|---|
| SiNVICT | v1.0 | Raw-reads-based | C++ & Python | Poisson model | 0.5 |
| outLyzer | v3 | Python | Noise level estimation | 1 | |
| Pisces | v5.3.0.0 | C# | Poisson model | – | |
| LoFreq | v2.1.5 | C & Python & Shell | Bernoulli trial | 0.05 | |
| UMI-VarCal | v2.5.0 | UMI-based | Python | Poisson test | 0.3 |
| smCounter2 | v2 | Python & R | Binomial distribution | 0.5 | |
| DeepSNVMiner | v1 | Perl & R | SAMtools calmd | 0.1 | |
| MAGERI | v1.1.1 | Java & Groovy | Beta-Binomial model | 0.1 |