-
Notifications
You must be signed in to change notification settings - Fork 69
-
Notifications
You must be signed in to change notification settings - Fork 69
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
RELAX with branch specific equilibrium frequencies #682
Comments
Dear @saurabh-mk, You can't assign equilibrium frequencies to branches; you can parameterize different substitution models that will induce different frequencies, and assign them to different branches. Where are you getting the frequencies for the branches that are being tested? Typically these frequencies are estimated by counts from the data. In order to implement this, you will need to edit the file Around line 200 in that file you will find the following statement, which defines the substitution model for the test branches. The equilibrium frequencies are passed in the
You could define a different set of frequencies by creating another variable
where Then pass Another issue that will occur at this point is that if you define a tree where models assigned to different branches have different equilibrium frequencies, HyPhy will require that you tell it which frequencies to use at the root of the tree (since the model is no longer in equilibrium). You can do so by changing the line
to
Regarding your question 2: yes, something like this can be implemented, but it requires considerably more work if you start with RELAX.bf, because you actually have to redefine the entire testing procedure, seeing how you now have two relaxation parameters (K_1 and K_2) and could runs tests like
... Best, |
Hi, I am sorry to report that I couldn't find these (or even similar) lines in the RELAX.bf file. I tried the current master as well as 2.2.6 and 2.3.4 versions. I found that lines similar to what you suggest already appear RELAX-Groups.bf, but I am not sure if they work the same way. For example, this is line 99 in RELAX-Groups.bf
What am I missing? Saurabh |
Hi, Would it be possible to say when this issue is likely to be answered? Meanwhile, I am wondering if you could make available the scripts use for non-homogeneous models in the RELAX paper (referred on pg 826, right column, first para of the paper)? Saurabh |
Dear @saurabh-mk, Apologies for the delay. Unfortunately, the original RELAX implementation from the paper is not compatible with the current release of HyPhy (stale code). I am attaching a version of Please place it inside
Currently, the analysis is set up to use the frequencies collected from sequences included in the test set, but you can look around line 300 to see other examples. Best, |
Hello! Thanks a lot for taking time to do this. Unfortunately, I get this error, after I run it as suggested (using the beta version)- "Expression appears to be incomplete/syntax error. Scope: 1, paretheses depth: 0, matrix scope: 0. Saurabh |
Dear @saurabh-mk, Hmm, can you check out the master branch (v2.3.11) and confirm the error? I am not able to replicate it. Best, |
Cant seem to get the 2.3.11 version. This
gives me this- HYPHY 2.3.1020180207beta(MP) for Linux on x86_64 |
Dear @saurabh-mk, 2.3.10 should also work. Best, |
Thanks for your responses. Really appreciate it. I get the same error with all the versions- "Expression appears to be incomplete/syntax error. Scope: 1, paretheses depth: 0, matrix scope: 0. Also on a different machine. I am at a loss how to trouble shoot/debug further. Can I provide any more details which may help? Saurabh |
Dear @saurabh-mk, Many apologies for this. I made a syntax error when commenting out example usage. See the attached file. Best, |
Hi, Thanks so much! Its now running, but it gives further errors during likelihood calculations
I have tried both analysis types- All and Minimal. They give the same errors. The likelihoods and the global estimates under the GTR model and the full codon models is exactly the same as the homogenous versions. So, I assume at that point its showing the output of fitting the homogenous model. I will try and dig into the RELAX-het.txt file to understand more. But do you have any idea of why the likelihood errors? I have attached the alignment file and the annotated tree. annotated_phylogeny.txt Saurabh |
Hello,
Sorry to revisit an old issue which remained unresolved. I hope it was ok to start a new one.
Can you guide me on how to accomplish this? What I have in mind is similar to Wertheim et al (MBE 2015) in the case of endosymbionts.
Thanks!
Saurabh Mahajan
The text was updated successfully, but these errors were encountered: