So you want single copy orthologs that also are only present in a subset of species? Maybe just run something like OrthoFinder and parse what you want from the output (i.e., selecting single copy alignments w/ members only from your subset)?

You don't need a tool per se. If you have for example have a data base where every row is an orthogroup and each column is the ID of every gene belonging to that family, you only need to identify those rows with a single not NA colum.

Using R-dplyr the solution would be something like:

orphan genes %

rowwise %>% summarise(

NA_per_orthogroup = sum(is.na(.))) %>%

filter(NA_per_orthogroup

>How would you go about finding these species-specific genes without simply checking every gene?

TLDR you wouldn't. The problem is, fundamentally, to check every gene and select only those that don't belong to an orthogroup. For well-studied eukaryotes, you can download a high-quality curated set of orthogroups from somewhere like https://www.orthodb.org Or if it is something non-model, you'll need to build them yourself using OrthoFinder, OrthoMCL or whatever.

Obviously, you don't need to check each gene manually, this is trivially automated in R, Python, bash or any other programing language you like.