https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1013470 If it holds what it promises, quite some potential together with #C07URFZF074

Thanks Daniel, already saw that! I am a bit reluctant to adopt it this early; first, because it's written in Python and will probably be quite slow (will need to benchmark though), and second because there is no conda env yet. Will keep an eye though, for an alternative way to create initial clusters 😉

Agree. I was also thinking of using the output from proteinfamilies as input to this tool.