-
|
Hi everyone, I'm trying to understand the substitution rate matrix inferred (in the .iqtree file). Rate matrix Q: A -1.248 0.3763 0.6966 0.1752 Best, |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
|
The rate matrix does not have units or a timescale. These are relative values, also known as instantaneous rates. You can convert these into probabilities of change along a branch of length https://revbayes.github.io/tutorials/ctmc/ As a side note (but an important one in this context) please note that in IQ-TREE, molecular dating is done by a separate program, LSD2, which is run after IQ-TREE finishes. LSD2 relies on the branch lengths and time constraints you supply. Thus, IQ-TREE itself does not do anything with your date information when calculating the Q matrix. Indeed, IQ-TREE is completely ignorant of any notion of absolute time. |
Beta Was this translation helpful? Give feedback.
The rate matrix does not have units or a timescale. These are relative values, also known as instantaneous rates.
You can convert these into probabilities of change along a branch of length
tby exponentiating this matrix. The RevBayes documentation is quite good on this (and there are videos too, though I haven't watched them!):https://revbayes.github.io/tutorials/ctmc/
As a side note (but an important one in this context) please note that in IQ-TREE, molecular dating is done by a separate program, LSD2, which is run after IQ-TREE finishes. LSD2 relies on the branch lengths and time constraints you supply. Thus, IQ-TREE itself does not do anything with your date information when calcula…