VCF to genome mapping

[PBM]

Hi All,

I wanted to know the right way to map the VCF back to genome.
The exercise is to guess from a given VCF file which genome it is made from.

I am fairly new to this field so any input will be fairly appreciated.

Thanks!

[kmcarr]

Hi All,

I wanted to know the right way to map the VCF back to genome.
The exercise is to guess from a given VCF file which genome it is made from.

I am fairly new to this field so any input will be fairly appreciated.

Thanks!

A VCF file represents already mapped data so "mapping a VCF to a genome" doesn't make sense. Further, there should be no guessing what the reference genome was. A properly formatted VCF should include ##reference and ##contig tags in the header which will unambiguously identify the species/genome version used for alignment and variant calling. If your file does not include this information, find the person who created the file and ask them for this information; then slap them for passing out poorly annotated files.

[id0]

A properly formatted VCF should include ##reference and ##contig tags in the header which will unambiguously identify the species/genome version used for alignment and variant calling. If your file does not include this information, find the person who created the file and ask them for this information; then slap them for passing out poorly annotated files.

I've seen VCF files that are publicly available as part of some major sequencing projects (in other words, not just being passed around a few collaborators) that do not have all that information. If you can't even trust those vetted VCFs, then it's a bit harsh to blame other creators.

[kmcarr]

I've seen VCF files that are publicly available as part of some major sequencing projects (in other words, not just being passed around a few collaborators) that do not have all that information. If you can't even trust those vetted VCFs, then it's a bit harsh to blame other creators.

Respectfully disagree. Absence of essential metadata makes the experimental data less than useful; we are, as PBM put it, left to "guess" what the data mean. Insisting that researchers provide the minimum metadata required to interpret and use their results is not harsh at all.

(And surely you recognized that my tongue was firmly planted in cheek wrt to the slapping comment.)