Reporting query k-mer positions?

[bananabenana]

Hi,

Thank you for developing this tool. Really enjoyed your preprint.

Query

I was wondering if there is a way to export the query k-mer positions for each reference position in the output vcf files? Or is there some way I can expose this as a temp file? The contig and position of the reference are present in the VCF files as expected, but not the query positional info.

Use case

I've been using kbo on assemblies in a pairwise manner, where each genome is used as a reference and a query for every other genome.
Example:

---Genome_dir
┕-----Genome_1.fasta
┕-----Genome_2.fasta
┕-----Genome_3.fasta

However, when I compare the same genomes as queries and references, I get different results:

Ref	Query	SNPs	Indels
Genome_1	Genome_2	79	10
Genome_2	Genome_1	86	11

I understand why this is occurring, but I'd really like to be able to see which reference positions these refer to, so I can remove common SNPs between both comparisons, and get the true number of coalesced pairwise SNPs (Genome_1 SNPs ∩ Genome_2 SNPs + Genome_1 SNPs + Genome_2 SNPs . This would be more accurate than taking the mean of them both I'd guess.

In these cases, the majority of these SNPs will overlap, but I cannot determine which ones do without the query positional information. Since I am using assemblies and not raw reads, this should be possible right?

Thanks

[tmaklin]

Hi Ben, thanks for the detailed report! I think this is indeed possible but it'll require some changes to the variant calling code, will have a look.