how to best annotate a merged pan-genome graph

[noahaus]

Hello,

I was curious to know how annotation is handled after 2 dbg graphs are merged using VariMerge? Is there an annotation tool (for bacteria) that the authors of this software would recommend alongside the update of pangenome graphs that might also save time? I'm interested in using VariMerge in some research very soon, but annotation is something I'm slightly concerned about.

best,
Noah A. Legall

[mmuggli]

I believe the term annotation means different things in different contexts. With the succinct colored de Bruijn graph, one can think of a row of the color matrix as annotating an edge in a de Bruijn graph that is the union of all the constituent samples used to create it. The VariMerge implementation has two separate programs. The first deals only with the succinct de Bruijn graphs and generates a union succinct de Bruijn graph of its two input graphs. It also writes out a "plan" file which is just a sequence of bytes with a very simple encoding (I don't remember exactly, but it was something like 0x1 = consumed and output k-mer from the left graph, 0x2 = consumed and output k-mer from the right graph, 0x3 = consumed a k-mer from each of the left and right graphs and output a single instance). The bytes indicate how the k-mers in the two inputs were merged and thus how to merge the color matrices. If one has their own additional annotation stored in the same order as the edges in the succinct de Bruijn graph, one could use the output plan file to merge their own annotations given a suitable program that knows how to read and write such custom annotations and how to combine annotations for k-mers that exist in both inputs.