Extreme OG size variation - definition

[Dayana-Turquetti]

Hi!

I am currently using CAFE v5.0 to analyze gene family evolution in Fabaceae species. These lineages often present relatively large gene families, which makes the preprocessing step particularly relevant.

In the documentation, it is mentioned that CAFE includes a built-in preprocessing step that filters out orthogroups (OGs) with extreme variation in gene counts. However, I could not find a precise definition of what is considered “extreme OG size variation” in this context.

Could you please clarify how this threshold is determined in CAFE v5.0? Is it based on a specific numerical cutoff, or is it handled adaptively depending on the data?

Thank you very much for your time and for making this tool available.

Best regards,
Dayana Turquetti

[hahnlab-user]

Hi Dayana,

There is no one size cutoff that CAFE uses, since it also depends on the depth of your phylogenetic tree. The best way to figure out the cutoff to use for your own dataset is to rank families by the variation in size among tips, and then to remove families from the top of the list (if the problem is that CAFE will not run).

matt