Dating with a fixed tree

[Daniel Vasquez Restrepo]

Hi everyone,

I'm here looking for a bit of help. I have a Maximum Likelihood topology that I would like to calibrate using fossil information. Since it's a relatively large tree, I was advised to infer the topology first and then perform the dating using RevBayes. I found this tutorial: "Dating with Relative Constraints", which I was able to follow up to step 2, where I'm unsure how to proceed.

From what I understand, the example suggests that it is possible to date an existing topology even if we don’t have the true timetree. However, the example focuses on doing this by comparing it to a known timetree, which I don’t have.

Any ideas on how I could modify the code? I already have a tree with inferred branch lengths, the posterior distribution of branch lengths (step 1a), and the mean and variance of those branch lengths (step 1b). I can define a reasonable prior for the root manually, and I also have my fossil calibrations. The only thing limiting me right now is the lack of a known timetree.

Daniel

[Ziv Lieberman]

Hi Daniel,

I may be misunderstanding some combination of your goal and what the tutorial does, but on my reading it seems that the comparison to the known timetree is only part of the tutorial to demonstrate the accuracy of the method. It doesn't seem to appear anywhere in the inference itself, only a posteriori analysis of results.

Secondly, are you sure that you need to be following this approach, rather than, e.g., this one?

Cheers,

Ziv