Hi,
I want to merge vcf files of same format from multiple samples using vcf-tools.But I want to know which method among vcf-merge or vcf-concat should I use for viewing results .Sample name does not matter in results.
This is the current format of each vcf file.
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT SampleName
Thanks
You would want vcf-merge. Concatenating means linking together like in a chain (so end-to-end).
I used vcf-merge, after sorting,bgzip and using tabix.Everything went well.But while merging I am getting this error below:My sample names are Clk51764 and Clk51769 as I am just testing the merge on 2 samples.
Using column name 'Clk51764' for 51764.sort.vcf.gz:Clk51764
Using column name 'Clk51769' for 51769.sort.vcf.gz:Clk51769
Could not determine the ploidy (nals=2, nvals=10). (TODO: ploidy bigger than 2)
6
at /vcftools_0.1.12/perl//Vcf.pm line 172
Vcf::throw('Vcf4_1=HASH(0x28a4e00)', 'Could not determine the ploidy (nals=2, nvals=10). (TODO: plo...', 6) called at /vcftools_0.1.12/perl//Vcf.pm line 2404
VcfReader::guess_ploidy('Vcf4_1=HASH(0x28a4e00)', 2, 10) called at /vcftools_0.1.12/perl//Vcf.pm line 1760
VcfReader::parse_AGtags('Vcf4_1=HASH(0x28a4e00)', 'HASH(0x2909ab8)') called at /vcftools_0.1.12/bin/vcf-merge line 464
main::merge_vcf_files('HASH(0x289cc20)') called at /vcftools_0.1.12/bin/vcf-merge line 12
If you can disregard the sample name (i.e. you're only interested in whether the SNPs are present, not where they came from), you can cut the first 9 columns, concatenate them together, and then run uniq over the resulting file.
for i in `ls *vcf`; do cut -f1-9 $i | grep -vE "^#" >> fullfile.vcf; grep ^# $i>uniqfile.vcf; done; sort fullfile.vcf | uniq > uniqfile.vcfNot tested so syntax might not be exact, but something like that will serve.
Edit: trying to make the code properly formatted