Open jun3234 opened 3 years ago
That message is saying that the REF allele in your vcf doesn't match your FASTA, apparently due to an off-by-1 error in your VCF.
The full error message you got should be:
error:[vg::Constructor] Variant/reference sequence mismatch: GTCAACTGGCTGCTGGCGCAGTGGTAGCGCCAGCAGCCAGCCCTGCCTCCCTTTGTAAGGGGGCAGGGTTCGACACCCTATATGTATATATATCTATATATAATATAATATATCTATAATAGTAATTGTTGTCCCACCATTTATGTGGGTAGTTCTCCCACCCTCGTGTGGGTTGTTTCTAGTTCTCTCCCAATTGGGAGTAGTCTTCTAAATGTCTATCTCGGTTGGTCCCGGCTCTGTCCCAATTTTGTACTCTTTCT vs pos: 295261: "<what it expects given your fasta>"
Variant: Chr1 295261 17 GTCAACTGGCTGCTGGCGCAGTGGTAGCGCCAGCAGCCAGCCCTGCCTCCCTTTGTAAGGGGGCAGGGTTCGACACCCTATATGTATATATATCTATATATAATATAATATATCTATAATAGTAATTGTTGTCCCACCATTTATGTGGGTAGTTCTCCCACCCTCGTGTGGGTTGTTTCTAGTTCTCTCCCAATTGGGAGTAGTCTTCTAAATGTCTATCTCGGTTGGTCCCGGCTCTGTCCCAATTTTGTACTCTTTCT N 0 PASSCHR2=Chr1;CIEND=0,0;CIPOS=0,0;END=295521;STRANDS=+-;SUPP=2;SUPP_VEC=11;SVLEN=-260;SVMETHOD=SURVIVOR1.0.7;SVTYPE=DEL
zero ind: 295260 1-indexed: 295261
You can check maually by running `samtools faidx
There's probably a better way, but you can check your vcf is valid by running it through bcftools norm -f
-- it should give you a similar error.
VCF file are 1-based format.
I havn't changed VCF since it's called with sniffles
.
It's weird that there is off-by-1 error in VCF file.
thanks !
I want to construct a graph from reference and SV vcf called with
sniffles
.I just keep one variant record as follow:
And used this command,
vg construct -r Lchinesis_genome.Chr.fasta -v test.vcf -S
. Then, it crashed with this info:If I was altered coordinate from
295261
to295262
, it would run successfully.Could you tell me details about SV vcf file format or any helpful links. Thank you in advance !