We need to review the step where we convert Graphtyper VCF outputs files to tabular format using vcf-to-tab

[masudermann]

Description of the bug

I've mentioned this before, but I don't think vcf-to-tab from VCFtools was designed to work with Graphtyper-formatted vcf files.

It works ok when dealing with haploid organisms where you are just filtering for homozygous SNPs and have haploid calls, but it in the conversion process to tabular format, output tables don't have any diploid calls (see screenshots below).

As a result, it also collapses any heterozygous SNPs.

I was reminded of this issue after Camilo used this tool with VCF files produced using GATK, and interestingly, all diploid calls were retained, as were heterozygous SNPs.

In the conversion to multi-fasta alignment, he then saw the expected ambiguity codes.

Let's discuss this more, but I think we may need to select another tool convert the vcf to tabular format, before creating the SNP-multifasta alignment files that are inputs into poppr and when making the SNP tree.

Finally, as I look at the VCF ouputs:

1. In the current iteration of the pipeline, during variant filtering, there is a line where only homozygous SNPs are retained regardless of ploidy. We should discuss the implications of this.

2. I think another flag needs to be added to retain only biallelic SNPs (if that is the goal).

Command used and terminal output

No response

Relevant files

Here is a snapshot of a bacterial output file generated using Graphtyper -> vcf-to-tab as part of a pipeline run

#CHROM POS REF GCF_002251695_1_ralstonia_solanacearum1 GCF_002251695_1_ralstonia_solanacearum2
NZ_NCTK01000001.1 28751 T G/ G/
NZ_NCTK01000001.1 28755 C T/ T/
NZ_NCTK01000001.1 28757 G A/ A/
NZ_NCTK01000001.1 28763 C G/ G/
NZ_NCTK01000001.1 28776 A G/ G/
NZ_NCTK01000001.1 28791 A G/ G/
NZ_NCTK01000001.1 28793 G T/ T/
NZ_NCTK01000001.1 28817 C C/ C/
NZ_NCTK01000001.1 28830 A A/ A/
NZ_NCTK01000001.1 28835 T T/ C/
NZ_NCTK01000001.1 28842 A G/ A/
NZ_NCTK01000001.1 28847 C A/ A/
NZ_NCTK01000001.1 28877 C T/ T/
NZ_NCTK01000001.1 28892 T C/ C/
NZ_NCTK01000001.1 28904 GA G/ GA/
NZ_NCTK01000001.1 28983 A G/ G/
NZ_NCTK01000001.1 29026 T C/ C/
NZ_NCTK01000001.1 29049 G G/ C/

---

Here is an oomycete output file generating using Graphtyoper -> vcf-to-tab (as part of same run)

#CHROM POS REF GCA_001466705_2_phytophthora_palmivora2
LATX02000001.1 1873814 G A/
LATX02000001.1 2111531 C T/
LATX02000007.1 1838 T C/
LATX02000012.1 104622 C ./
LATX02000013.1 24767 C T/
LATX02000013.1 665680 T G/
LATX02000017.1 2844 G A/
LATX02000017.1 55049 G A/
LATX02000020.1 52018 A G/
LATX02000020.1 52031 T C/
LATX02000024.1 498059 T A/
LATX02000046.1 457792 A ./
LATX02000046.1 457859 G T/
LATX02000046.1 457862 T C/
LATX02000046.1 457906 A C/
LATX02000046.1 457936 C ./
LATX02000046.1 457998 T C/
LATX02000057.1 373757 G A/
LATX02000057.1 406057 G C/
LATX02000057.1 442436 T C/
LATX02000078.1 731901 T C/

System information

No response

[cahuparo]

I agree with @masudermann! To provide more context here is the tab file produced by vcf-to-tab using an original gatk vcf file.