Build pipeline uses only ref VNTRs or all assemblies?

[ASLeonard]

Hi, I may be reading something wrong, but in the build pipeline, the config file seems to only take in VNTRs discovered from the reference (here), although the README seems ambiguous if it is ref-only (like tr.good.bed) or all assemblies.

Running danbing-tk build

• Required inputs:
• haplotype-resolved assemblies (FASTA)
• matched SRS data (BAM; optional)
• reference genome (major chromosomes only without minor contigs)
• tandem repeat regions (BED; available on release page or user-defined)

while from the manuscript methods definitely sounds like all assemblies are used TRF37 v4.09 (option: 2 7 7 80 10 50 500 -f -d -h) was used to roughly annotate the SSR regions of five PacBio assemblies (AK1, HG00514, HG00733, NA19240, NA24385).

Could you clarify if the build pipeline uses VTNRs discovered from each assembly or only the reference, and and how to include them if so?

Thanks, Alex

[joyeuxnoel8]

Hi Alex,

Thanks for the question. The idea of the pipeline is to take a set of VNTR regions on the reference coordinate, map regions from reference to each assembly, and then summarize the orthology mapping. So to answer your question, yes, all VNTRs from all assemblies are used but only the reference coordinates are required.

-Tony

[ASLeonard]

Great, thanks for clarifying Tony!