VarScan somatic indel caller working incorrectly?

[ElenaN]

Hi
I am seeing a seemingly incorrect somatic indel call using VarScan somatic.
The VarScan output line is
chr2 178883096 T +TAAC 11 0 0% T 6 2 25% */+TAAC Somatic 1 0.16374269 6 0 2 0 9 2 0 0

i.e. normal_reads2 = 0, yet the alignments and the pileup file show that the normal count is > 0. From the pileup file:
chr2 178883096 T 20 ...,...,....+4TAAC.+4TAAC.+3AAC.+3AAC.+4TAACC.+4TAAC.^KC ;;DHADDGEHH&1!!4!4.! 10 .......+4TAAC.+4TAACC. CDEEFF15!.

The parameters are:
samtools mpileup -f /home/enabieva/hg19/ucsc.hg19.fasta -A -E -q 1 $NORM $TUMOR >$OUT.pileup
java -jar ~/VarScan.v2.3.5.jar somatic $OUT.pileup $OUT.vs --mpileup 1 --tumor-purity 0.95

The VarScan version is 2.3.5
Anyone know what's going on?

Thanks.

[rnahar]

Hi
I am also getting some false somatic indel calls using Varscan 2.3.6.
For example
Indel 1 output
2 101582245 A -GCCCAGGCATGGGG 35 0 0% A 36 18 33.33% */-GCCCAGGCATGGGG Somatic 1.0 3.197076117906895E-5 1 351 17 2 33 0 0
However the normal pileup shows that the indel is present also in the normal sample
Indel 1 normal pileup
2 101582245 A 65 ,$,$,,,,-14gcccaggcatgggg,,,,.,,,,,-14gcccaggcatgggg,,,-14gcccaggcatgggg,-14gcccaggcatgggg,-14gcccaggcatgggg,-14gcccaggcatgggg,,-14gcccaggcatgggg,.,,-14gcccaggcatgggg,-14gcccaggcatgggg,,-14gcccaggcatgggg,-14gcccaggcatgggg,,,-14gcccaggcatgggg,-14gcccaggcatgggg,-14gcccaggcatgggg,-14gcccaggcatgggg,-14gcccaggcatgggg,-14gcccaggcatgggg,,,,,,-14gcccaggcatgggg,-14gcccaggcatgggg,,-14gcccaggcatgggg,-14gcccaggcatgggg,,-14gcccaggcatgggg,-14gcccaggcatgggg,,-14gcccaggcatgggg,-14gcccaggcatgggg,-14gcccaggcatgggg,,-14gcccaggcatgggg,-14gcccaggcatgggg,,-14gcccaggcatgggg,,-14gcccaggcatgggg, ;?@??????>;????1??1111?1?=>11@11>?111111??@?;11?11?11?111?11?1?1?

Indel 2 output
17 41121209 G -GC 20 0 0% G 22 7 24.14% */-GC Somatic 1.0 0.018169608718841608 9 13 5 2 1010 0 0
Again this deletion is present in the normal sample too.
Indel 2 normal pileup
17 41121209 G 26 ,$,-2gc.-2GC,-2gc,-2gc,,.....,.,-2gc,,,-2gc...,,,,. 8144422>;<>59<.:8.>>>9988
Sanger sequencing also confirmed this to be a germline indel.

There are a few more Germline indels being called as somatic which I had verified using Sanger sequencing. I had posted on the Varscan forum too more than a month ago but no response.
Can any expert reply on this - this is essential due to the lack of reliable somatic indel callers. Not getting proper support for Varscan. Let me know if more information is needed.
Thanks