Table 1.
Validation of the mutations within the targeted regions (76 genes) of the Illumina deep-sequencing data from 16 LUSC patients
| Mutation set | Non-somatic | Somatic |
|---|---|---|
| None |
4292 |
9 |
| Caller A only |
3 |
4 |
| Caller B only |
1 |
2 |
| Caller C only |
2 |
0 |
| Caller D only |
6 |
5 |
| Caller A and C |
0 |
6 |
| Caller A and D |
2 |
3 |
| Caller B and C |
0 |
2 |
| Caller B and D |
0 |
2 |
| Caller C and D |
0 |
5 |
| All but Caller D |
0 |
3 |
| All but Caller C |
0 |
4 |
| All but Caller B |
3 |
15 |
| All but Caller A |
0 |
13 |
| All callers | 0 | 57 |
The variants included in any mutation output (VCF) file were divided into mutation sets based on the detection status of the four callers (rows). For each variant, validation status (columns) was determined as ‘somatic’ if the deep-sequencing data shows the signature of a somatic mutation: the vaf in the tumor sample is > 10% and the vaf in the normal sample is < 2%; otherwise, it is as ‘non-somatic’. Variants that have less than 6 reads in the original exome-seq data were discarded. One mutation was such a case.