Regarding PAML output

[Priya Solaiman]

Hello All,

I'm trying to find the dn/ds ratio between 4 species, where 2 are monogamous and 2 are polygamous. I hypothesize that polygamous species will show positive selection, hence rapid evolution of genes compared to monogamous species. And I ran the codeml model 0, model 2, and model 2a. Now I'm confused about how to do the LRT test to find a better fit, and also, I'm worried that I might have extracted different dn/ds output from model 2. 

Also, I'm thinking I should have a different alternative model.

Can anyone give me some time to look at my analyses and point out if I made a mistake? Codeml is confusing to me. I appreciate your help.

[Priya Solaiman]

I have attached two screenshots of model 0 and model 2 ctl and tree files. My monogamous species are (hz & erec), and polygamous species are (ss& st).

[Sandra AC]

Hi there,

Thanks for your message! 

We wrote a protocol paper that shows how to use CODEML to test for positive selection (Álvarez-Carretero et al., 2023). The protocol is supposed to be read alongside the GitHub repository `positive-selection` -- you shall find in-house R scripts to carry out LRTs in this repository too.

Please note that you may want to go through the protocol aforementioned before running CODEML so that you can get familiar with the program, the control file, and how to format your input data. For more details on how to format the data for PAML programs, you can always visit the PAML Wiki (section "Data formatting").

All the best,
Sandra