build graph genome via cactus based on two genome

[Morriyaty]

Hi,

I only have two genome and I want to build a graph via cacuts. But a tree is needed for cactus. How can I construct tree based on two genome (maybe I cann't build tree contained two genome)?

Thanks : ) Yinjia

[Morriyaty]

Hi，

I have another question about input genome file. For each genome, it contained a lot chromosomes and some contigs. I intend to run cactus with split chromosomes. How can I handle this ?

Thanks. Yinjia

[glennhickey]

The tree is optional.

For splitting my chromosome, you can either do it yourself (ex with minimap2), or if you start with a minigraph, run cactus-graphmap-split as described here: https://github.com/ComparativeGenomicsToolkit/cactus/blob/master/doc/pangenome.md

Note that that document is still under construction

[Morriyaty]

Our genomes come from four different species. According to the "A cattle graph genome incorporating global breed diversity (:https://doi.org/10.1101/2021.06.23.449389)", they bulid a graph genome by soft-masked genome and a tree file. I intend to do my own work like this.

I have two questions about this:

1. What software do you recommend for building tree? (ex mashtree)
2. Can genome-wide tree representing each chromosoomes?

Thanks a lot. Yinjia

[glennhickey]

If you only have four genomes, I don't think you need to split by chromosome (unless you really don't have much memory). You can just send the species along with a tree into cactus.

For building the tree, again, with only four genomes there are only a few trees possible. mashtree is probably fine. You can also look at raxml if you have some gene annotations.