Example BAM file and output file are available?

[biotech25]

Hi Jeremiah,

I wonder if you provide example BAM files and output file (.vcf) that I can test svaba and replicate the result. It is because I like to test whether I am running svaba correctly or not.

I downloaded BAM files from ICGC, and ran svaba. Please see the attached files that I got from svaba running;

1. DO6350.svaba.somatic.sv.vcf.txt - This is filtered file.
2. DO6350.svaba.unfiltered.somatic.sv.vcf.txt - This is unfiltered file.

The FILTERED file is empty. All of variant callings have low quality in the UNFILTERED vcf file. I didn't use any parameter in svaba running. Please see the command I used below;

svaba run -t 474e3b36be9aba5e893234f2d4891e6e.bam -n 3a27910d544adbeab95b2b73c20ab6e5.bam
bwa index ./BWA_DNA_RefSeq_Fasta_GRCh37/Homo_sapiens.GRCh37.75.dna.primary_assembly.fa
-G ./BWA_DNA_RefSeq_Fasta_GRCh37/Homo_sapiens.GRCh37.75.dna.primary_assembly.fa
-p 16 -D ./DBSNP_knownIndel/dbsnp_indel.vcf
-a DO6350

I don't think the BAM files I used have problem or wrong file format. Therefore, I like to test svaba with example bam files and see whether I can replicate the result, and see whether I can get FILTERED (high quality) variant callings. Could you help?

Sanghoon

DO6350.svaba.somatic.sv.vcf.txt DO6350.svaba.unfiltered.somatic.sv.vcf.txt

[drvenki]

Hi Sanghoon, I have been using svaba for a while now and I think I understand why this issue occurs. svaba takes into consideration only those events which have good coverage. Those with less supporting reads do not survive the filtration. It is common for me as well to find empty vcfs. For those variants that you think you might want to see, they will all be in the unfiltered file. You could relax the stringency cutoffs and very well produce a new vcf. For all such steps, the bps.txt.gz will be the mother file. Hope this helps! Venkatesh Chellappa (Venki)

[biotech25]

Thank for your answer Venkatesh. I am going to follow your advice and try again.

Last time, I found that I didn't index BAM file, and just ran SvABA. Now, I am re-running SvABA after indexing BAM file. I hope this can make difference and output VCF file with some detected SVs. If I still can't get any SV in vcf file, I will try your advice.