Pindel: Reference counting (Short Insertion)

[renatopuga]

I have a problem with counting sequences in reference to the Pindel software.

I am analyzing exome data (tumor vs. normal) and found a Short Insertion (SI) using Pindel.

One inclusion of a base (A) CTC[A]ATCA.
Total reads: 81 reads (Ref) and 83 reads (Alt) = 164 (DP), counting in tumor samples. No variant in the normal sample.

So, I used the IGV to visualize the integration (knowing that the counting of IGV can be different).
The IGV does not show sequences with insertion.

After that, I used some softwares to count/extract sequences in the insert region (counting down).

softwares        Ref	Alt	DP
--------------------------------------
pindel		81	83	164
atlas indel2	6	92	98
indelocator	6	92	98
bam-readcount	95	91	186
samtools view	8	92	100

My question is: How can I confirm that the sequences are equal to Ref only the Pindel and bam-ReadCount listed?

I know Pindel realigns the sequences, but the VAF in Pindel is 50% and the other ~ 95%

- illumina HiSeq 2000
- Exome - Pairend - 150x

- Bwa Version: 0.7.5a-R405
- Human_g1k_v37.fasta reference

- Pindel version 0.2.5a1, July 23 2013.
- Atlas Indel2 v1.4.3-R158
- IndelGenotyper.36.3336
- Bam-0.4.6-ReadCount
- Samtools-0.1.19 44428cd

My Best.

[renatopuga]

The supported number of reads is the same as normal reads. Because the count for Insertion Sort (SI) is initialized based on the next

Reference
123456789
AAAT-AAAA

AAATCAAAA
AAATCAAAA
AAATCAAAA
AAAT-AAAA
AAAT-AAAA

One insertion (C) at position 4 (T): Ref(T) = 6 Alt(C) = 3.

Pindel sample:
NumSupSamples 1 tumor 6 6 3 3 0 0

DP = Ref + Alt (depth total)

The six reads are equal to Ref (T), where three of these reads contains the insert (C). My DP = 9, but the correct is: DP = 6.