hal2maf into vcf

[tzaquin]

Hello,
I was using cactus-progressive to create a pangenome from multiple species. Now that I want to analyze the variations between samples, I find myself a bit stuck.
I used the cactus-hal2maf using a reference to an ancestor. Next, I tried to get vcf file using maf2vcf. Yet I'm getting the following error:

ERROR: Couldn't find a header line (must start with Hugo_Symbol, Chromosome, or Tumor_Sample_Barcode): cactus.maf

and this is the maf file:

##maf version=1 scoring=N/A

a
s       Anc3.Anc3refChr0        0       81      +       2129    G----------ctaaccctaaccct--aaccctaaccctaaccc-taaccccaaaccctaac-cctaccccaaaccctaacctaaaccctaaccc
s       Thunnus_orientalis.scaffold1    0       82      +       34771766        G----------ctaaccctaaccct--aaccctaaccctaaccc-taacacctaaccctaacgcctacccaaaaccctaacctaaaccctaaccc
s       Thunnus_thynnus.scaffold1       0       94      +       34640894        TACCCTAAACCCaaaccctaaccctcaaaccctaaccctaacccctaaccccaaaccctaac-cctacccctacccctaacccaaaccctaaccc

a
s       Anc3.Anc3refChr0        81      408     +       2129    taaccctaa-ccctaaccctaaccctaaccctaaccctaaccctaaccctaaccctaaccctaaccctaaccctaccccaaacccta-ccctaaccctaaccctaacccgaacccaac---cctaaccc-taaccctaacccgaaccctaacccta-ccctaacccaaaccctacccctaaccctaccctaaccctaaccctaaccctaaccctaaccctaaccctaaccc-taaccctaa-ccctaaccctaaccctacccctaaccctaaccc-taaccctaaccctaaccctaaccc------------tatccctaaccctaaccctaaccc--aaccctaacccaaccctaaccctaaccctaaccctaaccctaaccctaaccctca---------ccctaaccctaaccctaaccctaaccctaaccctaaccc
s       Thunnus_orientalis.scaffold1    82      409     +       34771766        taaccctaa-cccgaaccctaacccgaaccctaaccctaaccctaaccctaaccctaaccctaaccctaaccctaacccgaacccta-cccgaaccctaaccctaacccgaacccaac---cctaaccc-taaccctaacccgaaccctaaccctaaccctaaccctaaccctaaccctaacccgaacctaaccctaaccctaaccctaaccctaaccctaaccctaaccc-taaccctaa-ccctaaccctaaccctaaccctaaccctaaccc-taaccctaaccctaaccctaaccc------------tatccctaaccctaaccctaccct--aaccctaacctaaccctaaccctaaccctaaccctaaccctaaccctaaccctaa---------ccctaaccctaaccctaaccctaaccctaaccctaaccc

I would much appreciate any help. Maybe you could recommend a different method to getting the vcf file?
Thank you

[glennhickey]

The MAF format output by cactus-hal2maf is described here, where there is no mention of Hugo_Symbol etc. I guess you will need to consult with the authors or documentation of maf2vcf and add in the appropriate header yourself.

If you want a VCF directly from Cactus, you need to use the pangenome pipeline, but that only works for samples from the same or very closely related species (I'd also argue the VCF format itself is mostly suited for this type of data)

[tzaquin]

Thank you for your response.
When you are saying "closely related species", what might be a distance cut off? For example, I'm working on species from the same genus which separated around 3MYA.

Thank you

[glennhickey]

I'd say 3MYA is definitely worth trying out with the pangenome pipeline.