Question: Combine Single Sample Vcf To Multi Sample Vcf
5
gravatar for William
6.7 years ago by
William4.5k
Europe
William4.5k wrote:

Is it possible to combine single sample VCF files to multi sample VCF files? The samples are different strains from the same species, mapped against the same reference.

The single sample vcf files have been made with GATK and the multi sample vcf file should preferentially also be compatible with GATK for downstream processing.

We chose not to do multi-sample calling in GATK because we don't want the genotype calls and their quality from one sample to be influenced by the genotype calls and their quality from other samples (which GATK does by default when doing multi-sample calling ).

We do however want to to process the samples the and their genotypes as one big matrix.

vcf gatk snp • 11k views
ADD COMMENTlink modified 6.5 years ago by alexej.knaus120 • written 6.7 years ago by William4.5k
6
gravatar for swbarnes2
6.7 years ago by
swbarnes27.4k
United States
swbarnes27.4k wrote:

The problem is that unless your vcf's have a line for every single locus, samples that match consensus at a position will not have an entry for that position, so if you have a SNP in one sample at a given position, other samples will have no vcf entry for that position, and you won't know if your other samples really match consensus there, or if they have low coverage there, and a variant couldn't be safely called.

That's why calling them all together is preferable. If one variant is iffy in one sample, the fact that that variant exists in other samples makes it more likely that the the iffy variant is real.

ADD COMMENTlink modified 13 months ago by RamRS25k • written 6.7 years ago by swbarnes27.4k
1

I did use EMIT_ALL_CONFIDENT_SITES for the output mode of the unified genotyper. So each sample vcf also has reference calls.

ADD REPLYlink modified 13 months ago by RamRS25k • written 6.7 years ago by William4.5k
5
gravatar for Pierre Lindenbaum
6.7 years ago by
France/Nantes/Institut du Thorax - INSERM UMR1087
Pierre Lindenbaum126k wrote:

use vcf-merge

ADD COMMENTlink modified 13 months ago by RamRS25k • written 6.7 years ago by Pierre Lindenbaum126k
1

Documentation migrated to https://vcftools.github.io/perl_module.html#vcf-merge

ADD REPLYlink modified 13 months ago by RamRS25k • written 4.1 years ago by setempler20
0
gravatar for alexej.knaus
6.5 years ago by
alexej.knaus120
Berlin
alexej.knaus120 wrote:

You can also try Gene-Talk.de, upload the single files, and create a collection from all. afterwards you can filter and reduce your data to clinically relevant variants.

Up to now GeneTalk supports only human hg19 referenced VCF files.

ADD COMMENTlink modified 13 months ago by RamRS25k • written 6.5 years ago by alexej.knaus120
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 2075 users visited in the last hour