POS, SVLEN, and END are seemingly discordant in VCF output

[RyanVidegar-Laird]

Hi, thanks for all of your work on this tool!

I've ran xTea (with defaults) on 4 short-read WGS samples, and am a bit confused why the POS, END, and SVLEN values don't seem to align in the VCF output. I would expect END = POS + SVLEN, yet it doesn't across any of my samples for Alu or L1 SVs. Is this an error? I'm new to working with SVs, so perhaps it's my misunderstanding.

Small output example: awk '!/orphan/' ./xtea/out/sample-01_ALU.vcf | bcftools query -f'[%CHROM\t%POS\t%INFO/SVLEN\t%END\n]' - | shuf -n 5 | awk '{$5 = $4-$2}1' | column -t

CHR	POS	SVLEN	END	END-POS
chr6	49594200	269	49594216	16
chr9	96301414	276	96301428	14
chr8	19801418	292	19801427	9
chr2	64936633	274	64936649	16
chr8	127025867	378	127025879	12

[simoncchu]

This is insertion, which means they are absent from the genome. Here, SVLEN is the insertion length. In general, we think one insertion will have one breakpoint on the genome, however for TE insertions the two strands usually do not break at the exact same location (you can search for target-site-duplication in L1 retrotransposon to understand more). Thus, there are two breakpoints reported here (POS and END). Hope this helps.