Entering edit mode
11.7 years ago
Pierre Lindenbaum
161k
I've indexed a set of vcf with tabix.
$ gunzip -c A.vcf.bgz
##fileformat=VCFv4.1
(...)
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT A
chr1 10109 . A T 101.0 . DP=101;VDB=0.0164;AF1=0.5;AC1=1;DP4=27,19,13,23;MQ=25;FQ=104;PV4=0.049,9.6e-06,0.056,0.15 GT:PL:GQ 0/1:131,0,233:99
chr1 10177 . ACC ACCC 70.5 . INDEL;DP=56;VDB=0.0099;AF1=1;AC1=2;DP4=0,0,0,27;MQ=21;FQ=-116 GT:PL:GQ 1/1:111,81,0:99
chr1 10179 . C CN 70.5 . INDEL;DP=59;VDB=0.0140;AF1=1;AC1=2;DP4=0,0,0,27;MQ=21;FQ=-116 GT:PL:GQ 1/1:111,81,0:99
chr1 10230 . ACCCC ACCC 9.02 . INDEL;DP=50;VDB=0.0322;AF1=0.6672;AC1=1;DP4=0,2,0,8;MQ=20;FQ=-31.5;PV4=1,1,0.051,0.42 GT:PL:GQ 0/1:45,
0,3:7
$ tabix A.vcf.bgz chr1:10177-10178
chr1 10177 . ACC ACCC 70.5 . INDEL;DP=56;VDB=0.0099;AF1=1;AC1=2;DP4=0,0,0,27;MQ=21;FQ=-116 GT:PL:GQ 1/1:111,81,0:99
tabix and vcf-tools/bin are in my $PATH
I've set PERL5LIB to /path/to/vcftools0.1.9/lib/perl5/siteperl
when I try to merge my VCFs with VCFtools I get the following error:
$ vcf-merge A.vcf.bgz B.vcf.bgz
Could not parse the fileformat version string [ (#######large BINARY string###############)
], assuming VCFv4.1
Broken VCF header, no column names?
at /usr/local/package/vcftools_0.1.9/lib/perl5/site_perl/Vcf.pm line 177
Vcf::throw('Vcf4_1=HASH(0x1e04e200)', 'Broken VCF header, no column names?') called at /usr/local/package/vcftools_0.1.9/lib/perl5/site_perl/Vcf.pm line 868
VcfReader::_read_column_names('Vcf4_1=HASH(0x1e04e200)') called at /usr/local/package/vcftools_0.1.9/lib/perl5/site_perl/Vcf.pm line 610
VcfReader::parse_header('Vcf4_1=HASH(0x1e04e200)') called at /usr/local/package/vcftools_0.1.9/bin/vcf-merge line 168
main::init_cols('HASH(0x1e3d73d0)', 'Vcf4_1=HASH(0x1e04dc20)') called at /usr/local/package/vcftools_0.1.9/bin/vcf-merge line 264
main::merge_vcf_files('HASH(0x1e3d73d0)') called at /usr/local/package/vcftools_0.1.9/bin/vcf-merge line 12
what's wrong with my command ?
Pierre
I have the same problem as Pierre, just with a different tool. I also used bgzip, tabix and added the perlpath. Just my vcf file has a .gz and not a .bgz, but still I get the same error as Pierre. Any additional suggestions? Thanks Eva
Did you get any error messages while runing tabix? ; check if VCF is sorted or not? I generally follow this:
for F in *.vcf ; do vcf-sort $F | bgzip -c > ${F}.gz ; tabix -p vcf ${F}.gz ; doneHi I have the similar errors. Have you sorted it out ?
Thank in advance
goto
http://bioinfoconcepts.blogspot.in/2015/04/vcf-tools-error-broken-vcf-header-no.html?view=sidebar
You do realize this is a 2 year old post, right? And you're just pointing to a (not so wholesome) solution that you've written in a blog that is essentially the same as Brad's solution from two years ago. That's not really helpful.