How to compare VCF files called from different reference genomes?

Question

[Begonia_pavonina]

Hello everyone,

How to compare two VCF files called with the same software (FreeBayes), but with completely different reference genomes based on the same specie?

The question has been asked here already: C: Comparing VCFs constructed from mapping to different reference genomes

But it seems that LiftoverVcf is made to adjust VCF sequence called with similar genomes rather than with completely different reference sequences (with different headers).

Would you have any trick to suggest to compare these VCF files?

[Begonia_pavonina]

Thanks for your answer igor. I was planning to compare a VCF file mapped to a genome to target-capture VCFs mapped to the baits sequence.

Eventually and as you quote, it is easier to map the target-capture dataset to the same reference genome than the first VCF file. Thanks for the help!

[igor]

It's generally a good idea to map to the full genome, not just the target regions. Your target capture protocol does not capture just the baits.

Answer 1

[igor]

LiftoverVcf is made to adjust VCF sequence called with similar genomes rather than with completely different reference sequences (with different headers)

You can lift over even across species, such as between human and mouse.

Why not just generate the VCFs using the same reference?

[Begonia_pavonina]

Thanks again for your comment igor.

I have an annotated file with trusted variants from another specie (non-model species) I wish to compare to the variants I have generated to check them.

I thought I could not be done, but as you quote it may be possible. I will have to generate my own chain file. http://genomewiki.ucsc.edu/index.php/LiftOver_Howto

Has anyone tried this already?