Hi, Arndt pointed me towards your poster at SMBE last week, and the method looks interesting. I have a tree with some very long branches and would like to gauge whether the parameter estimates are likely to be problematic. Unfortunately, I think the output files from the version of IQ-TREE2 that I used may not work with the parsing code.
I found that patching -mfreq into modelAndFrequency was necessary to get the script to successfully run IQ-TREE version 2.1.3 on the subsequences/clades. When if multiplicity == 1 the variable M is undefined and causes an error. The string "Rate parameter R:" also doesn't appear in the *.iqtree files I get from amino acid alignments.
Could you confirm which version of IQ-TREE2 you are using?