Dear Vincenzo,
For ML estimates the edge length are the expected number of substitutions per
site, which depends on the product of rate and time. So you need either a rate
estimate or calibration points to estimate the time. If you divide the edge
length by the rate the edges should be
Dear Vincenzo.
If I understand your problem problem, you do not have any explicit
calibration points -- but you have a hypothesized average clock rate?
If so, then you can obtain an ultrametric tree from/ape::chronos/ for
any value of the smoothing parameter (/lambda/) and then simply re-scale
Dear Emmanuel,
Thank you very much for your response. I cannot see how to provide the
substitution rate to the phangorn::pml_bb() function, but I was looking at
the ape::node.dating() function and it appears that I could provide the
substitution rate to the "mu" argument and then set the
Dear Liam,
Thank you for your answer. Yes, I have a hypothesized average clock rate
and no explicit calibration points.
If I use ape::chronos() with default values it gives me a depth of 1 at the
root. So I suppose to rescale the branch lengths I just need to multiply
all of the branch lengths