svdb merge changes VCF header

[Thatguy027]

Hi there,

I merged four SV callers using svdb merge (delly, smoove, manta, tiddit). Then I tried running the merged VCF through vcfanno to annotate with genomic features and ran into an error. I checked on the vcfanno issues page and people had similar errors when attempting to annotate with malformed VCFs.

After looking at the header I noticed this line was different between the single caller and merged VCFs:

##INFO=<ID=END,Number=1,Type=String,Description="End of an intra-chromosomal variant">

My quick fix was to run this command on the VCF:

sed 's/ID=END,Number=1,Type=String/ID=END,Number=1,Type=Integer/g'

This resolved the issue with vcfanno.

I thought I'd let you know!

[J35P312]

Hello!
Thanks for your feedback, and nice to hear that you found a solution!
Have you checked the TIDDIT header? I believe the

##INFO=<ID=END,Number=1,Type=String,Description="End of an intra-chromosomal variant">

originates from an older version of TIDDIT (it looks very familiar to me ;P). I fixed it relatively recently, so if you download the latest TIDDIT it will be fixed. Alternatively you may change the order of vcf files when merging: svdb --merge --vcf manta.vcf tiddit.vcf delly.vcf smoove.vcf

This should give you first the manta vcf header (SVDB prints the union of each vcf header item, prioritised based on the order of the input vcf files).

Interesting, I have never heard about smoove, I will have a look!
Thanks
//Jesper