Table 1 Evaluation of INDELseek complex indel detection performance
From: INDELseek: detection of complex insertions and deletions from next-generation sequencing data
Dataset | Sample count and description | Sensitivity | Specificity |
|---|---|---|---|
Real NGS data | |||
1. Protein-coding and flanking regions from whole-genome sequencing (random fragments) | 1 (NA12878) | 100% | 100% |
160 putative complex indels | |||
26 negative control loci | |||
2. Hereditary breast and/or ovarian cancer panel (amplicons) | 239 | 100% | 100% |
3 positive samples (BRCA1 n = 1, BRCA2 n = 2) | |||
236 negative samples | |||
3. Myeloid neoplasm panel (amplicons) | 23 | 100% | 100% |
5 positive samples (CALR n = 4, JAK2 n = 1) | |||
18 negative samples (NA12878 and 17 healthy controls) | |||
Semi-simulated data by engineering mutations to real NGS data | |||
1. Whole-genome sequencing (random fragments) | 8671 collected from COSMIC and dbSNP | 93.7% | N/A |
2. Hereditary breast and/or ovarian cancer panel (amplicons) | 237 collected from COSMIC and dbSNP | 96.2% | N/A |
3. Myeloid neoplasm panel (amplicons) | 576 collected from COSMIC and dbSNP | 94.6% | N/A |