Question about filtering Diamond results

[Hannah1746]

Hello,

I am making a GENESPACE plot across 5 genomes that all shared a common whole genome duplication. I am getting some artifacts of synteny between the homoeologous chromosomes between the genomes:

One of the best examples in this photo is between Myx14 and My13 to Cac14 and Cac11 I am hoping that I could maybe have a way to filter the Diamond results in attempt to filter these artifacts out. I understand there is some internal filtering that GENESPACE is already doing but I can't seem to isolate where that step is in the process.

Ether way I would love to hear your thought on how to deal with this without just brut forcing it.

[jtlovell]

to be clear, you are using default specifications with ploidy = 1 and no outgroup? If so, I am surprised that you get so much subgenome pollution.

[Hannah1746]

Yes. I am runnning: gpar <- init_genespace( wd = wd, path2mcscanx = path2mcscanx, genomeIDs = c("MX", "RBS", "HN1", "MO1", "CC5"), ploidy = 1 )

and then:

out <- run_genespace(gpar, overwrite = T)

[jtlovell]

Interesting - the only time that I have seen this type of pattern is when the WGD happened just before the genomes diverged ... this can confuse things. I would say to try to increase the block size init_genespace(..., blkSize = 20) ... i that doesn't do it, reach out to me directly (bluesky = @jotlovell, email, etc.) and we'll get it figured out.

[Hannah1746]

This defiantly helped but there are still some artifacts.

[jtlovell]

yeah, I think the dup is likely close to the root of the tree.