chomo, cscores and consistency index

9 views
Skip to first unread message

Jakob Jilg

unread,
Mar 16, 2026, 6:55:20 PM (3 days ago) Mar 16
to TNT-Tree Analysis using New Technology
Hi all,
to measure homoplasy for individual characters, I want to calculate the consistency index (CI) by "mapping" the matrix onto the topology of a nucleotide tree.

I'd do it like this:
proc matrix.tnt;
proc nucleotide.treefile;
chomo;
cscores;

The problem is:
What exactly are chomo and cscores reporting? I couldn't find much documentation, only the help command in TNT itself.
chomo: show homoplasy (=extra steps) for tree
cscores: report length/fit for character(s) L on tree(s)

The wording is unclear to me. Does cscores report the length divided by fit? Or are length and fit synonyms?

I interpreted the command like this: chomo reports "extra steps" for the tree (=observed number of changes minus minimum possible number of changes), and cscores reports the observed number of changes (=steps) for the character on the tree. In that case, the minimum number of changes possible would be the value of cscores minus the value of chomo for that character.
If this is correct, i'd calculate CI for individual characters with their value for: (cscores - chomo)/cscores

Example: For one character, chomo reported 6 and cscores reported 8. (8-6)/8=0.25

I just wanted to know if my interpretation of the output by chomo and cscores is correct.

Best,
Jakob
Reply all
Reply to author
Forward
0 new messages