Corrupted dbsnp-vcf?

Question

[Filago]

Hello,

I have a vcf file, which I want to upload on the Sanger Imputation server. The following error occured:

--- Aborted Job ---
The input file sanity check failed, "bcftools norm -ce" exited with the following message:
Reference allele mismatch at X:3155141 .. REF_SEQ:'T' vs VCF:'G'

As suggested by the sanger website, I wanted to solve this issue with the bcftools +fixref command.

All my SNPs have dbsnp-IDs, so I downloaded the following file for reordering alleles: ftp://ftp.ncbi.nih.gov/snp/organisms/human_9606_b151_GRCh37p13/VCF/All_20180423.vcf.gz

When I now use the

bcftools +fixref broken.vcf -O z -o fixref.vcf -- -d -f /path/to/reference.fasta -i `All_20151104.vcf.gz`

command, the following error appears:

[E::bgzf_uncompress] Inflate operation failed: invalid distance too far back
[E::bgzf_read_block] Invalid BGZF header at offset 15203091877

It seems, that the All_20151104.vcf.gz file is corrupted. I also am not able to index it with bcftools. However, another operation (subsetting it to regions) works...

Does anyone know, how to solve this problem?

Best,

Andreas

[Pierre Lindenbaum]

hg19: chrX:3155141 is T

hg18: chrX:3155141 is G

aren't you mixing hg* builds ?

[Filago]

I think/hope not...everything should be hg 19... Might be a stupid question, but where can I quickly check this for some SNPs?

Answer 1

[Filago]

Ok I solved it! The problem was the following: During the preprocessing, I converted plink files to vcfs and assumed the first allele in the bim file to be the reference allele...however, it is not! I now solved it by using the --ref-from-fa command when making vcf files.