I'm trying to generate a GAF (with reference rather than node coordinates) which describes how an individual long-read maps to all the different components of a pangenome generated by minigraph-cactus . Initially I was looking at the following issue on the vg Github repo. The suggested solution isn't compatible with my situation (I think) because vg giraffe does not at present allow for long-read alignment. If I use the rGFA generated by the minigraph-cactus and minigraph for alignment, I do get the type of GAF with the limitation that I only see the primary reference coordinates. In the aforementioned Github issue there is a suggestion to use vg gbwt -Z --set-tag to add the other paths as reference into the gbz index, which seems to work as when I use vg convert to generate the GFA I can see the other paths listed as reference. However, this resultant file appears to be a GFA rather than rGFA which I think I would need to use in order to use minigraph to generate the type of GAF I need.
My question is then would anyone know if there is a better way of generating the type of GAF where the match of a given read to all references in the graph is provided in the sixth column? Or, is there a way to convert a .gbz into a rGFA?
