CNV benchmarking: Truvari & HG002 SVs v0.6 (GIAB)

Question

[emmanouil.a]

Hi,

I ran truvari using the "HG002_SVs_Tier1_v0.6.vcf.gz" (both as true and query), with "HG002_SVs_Tier1_v0.6.bed", and everything is working.

Looking the HG002_SVs_Tier1_v0.6.vcf.gz vcf has the bellow format

#CHROM  POS     ID      REF     ALT

1   15785403   someID  TAAAAAAAAAAAAAAAAAAAA   T

My query format is like

1    15785403        someID:LOSS:1:15785403-15785423  N       DEL

Is there any tool to convert my query format in something like the v0.6 format?

Thanks

[DBScan]

With which tool did you get your query format? To me this doesn't look like a valid VCF, but I may be wrong here.

[emmanouil.a]

Hi, Illumina Basespace DRAGEN, but I also have another pipeline that the format of "REF" and "ALT" is the same as in my query vcf. I do not remember if also DRAGEN-GATK gives the same format (I do not have an example in my hands now).

[DBScan]

I just checked with my DRAGEN VCFs and I have the same format. In their DRAGEN paper they used witty.er to do comparison between query and truth, see here https://www.biorxiv.org/content/10.1101/2024.01.02.573821v1.full.pdf, maybe you can find some more information there.