Using transcriptomic data, I have identified a suite of genes in my focal species that I hypothesise should be rapidly-evolving relative to the rest of this species' genome. I'd like to estimate ω (aka dN/dS) for each gene within this species, with a view to then testing whether the average ω of genes within my setof interest is greater than the genome-wide average ω for this species. My plan is to use OrthoFinder to generate orthologues between my species and others to which it is related, align these orthologues using PRANK, and then calculate ω using CODEML. I believe that this should be possible but I'm new to this type of genomics so do let me know if my entire plan is ill-conceived!
As I understand it, to generate ω for a species, one needs to be able to compare that species' genome to the genomes of multiple other species, ideally representing multiple different degrees of phylogenetic distance. For example, if I were estimating ω for the human genome, I'd want to compare to the genomes of our closest relatives (chimpanzees), and also to other more distant (but not too distant) relatives like gorillas and orangutans.
The specific problem I have is that my focal species represents an outgroup within its superfamily. Several other species in this clade have genome assemblies available, but they are all on 'the other side' of the phylogeny, so to speak: that is, the related species for which genomes are available are all much more closely related to one another than they are to my focal species (there are many other species in the clade that are more closely related to mine, but none seem to have genomes available). Other species' genomes are available outside this superfamily, of course, but they are all relatively distant relatives, and I have been unable to reliably generate orthologues between my species and these more distant relatives.
My question is: will it still be possible to reliably estimate ω for my species' genome, given that the other species to which I would be comparing it are all members of a single (separate) clade? This would be analogous to estimating ω for humans when the only available primate genomes are for several species of gibbons.
