Represent Precise Deletion In Vcf

Question

[Tomáš Beluský]

Hi, in my master thesis I have to implement tool for detecting genome variations. Now, I am learning how to represent structural variations in VCF file. But after reading example in http://www.1000genomes.org/wiki/Analysis/Variant%20Call%20Format/vcf-variant-call-format-version-41 I am little bit confused.

Example says:

#CHROM  POS ID REF ALT QUAL FILTER INFO FORMAT NA00001
1 2827693 . CCGTGGATGCGGGGACCCGCATCCCCTCTCCCTTCACAGCTGAGTGACCCACATCCCCTCTCCCCTCGCA C . PASS SVTYPE=DEL;END=2827680;BKPTID=Pindel_LCS_D1099159;HOMLEN=1;HOMSEQ=C;SVLEN=-66 GT:GQ 1/1:13.9

Length of reference sequence is 70bp (69bp if we remove first base which doesn't belong to deletion), but SVLEN here is 66bp and also if we subtract END with POS, we get -13. Shouldn't SVLEN be 69 and END position 2827762? Or maybe I'm missing someting here.

Thanks for clarifying me this.

[PoGibas]

Congratulations on having the same master thesis project as me! :-)

Answer 1

[bhandsaker]

Yes, I think this is definitely buggy. I believe this record is supposed to represent rs2376870 (1000 Genomes ID P1M0615101909), which on hg18 should look like:

1 2827694 P1_M_061510_1_909 CGTGGATGCGGGGACCCGCATCCCCTCTCCCTTCACAGCTGAGTGACCCACATCCCCTCTCCCCTCGCA C . . BKPTID=Pindel_LCS_D1099159;END=2827762;HOMLEN=1;HOMSEQ=G;SVLEN=-68;SVTYPE=DEL

I amended the spec, versions 4.1 and 4.2 (draft). I also fixed the typo on the next line where SVLEN should have been -205, not -105.

[Dan]

Can we move the docs to a public wiki?

Answer 2

[zam.iqbal.genome]

SVLEN is defined as length(alt) - length(ref), so looks to me like it should be 69, and that is a typo.

Answer 3

[Adam]

Probably best to ask on the vcf specification mailing list ([e-mail]), but I agree that it's probably a typo.