How to interpret the results for branch-site in ete-evol?

17 views
Skip to first unread message

Elisa Josefa González Ugalde

unread,
May 6, 2024, 1:59:23 AM5/6/24
to The ETE toolkit

Hello to the entire ete3 community!

I'm new here, I'm very sorry if this question has already been asked before; I searched and couldn't find an answer. I have a question about ete-evol, specifically about interpreting the results of branch-site analysis. I'm searching for genes and sites that are under positive selection in a whole-genome analysis, using only autosomal coding sites for a particular lineage (using koala as an outgroup). This is the tree I used for the analysis:

(((M3046 #1, M3048 #1), (M3047 #1, (M3049 #1, M3050 #1))), koala);

This is the command I used to generate the Branch-site results:

"ete3 evol -t /data1/lab1/elisa/def_ete3toolkit/CHR_AUTO/BRANCH_04/tres/tree.nwk --alg /data1/lab1/elisa/def_ete3toolkit/CHR_AUTO/NOKOALA_P5/${i}.fa --models bsA bsA1 --clear_all -v 3 -o /data1/lab1/elisa/def_ete3toolkit/CHR_AUTO/BRANCH_04/resultados_codeml/${i} --cpu 3

My question (actually, questions) are as follows:

  1. I'm not sure if in this case I should apply a correction for multiple hypotheses, since I only ran the models for a specific branch. Also, upon reviewing the PAML discussion group, it seems unnecessary if it is known a priori that there were indeed signs of selection in the studied lineage.
  2. I don't understand how to interpret the results of the sites in the Branch-site model. Finally, the value of the LRT statistic that is printed is for the branch, not for the sites; does this mean that the genes indicated as significant contain at least one site under positive selection? If so, is it safe to asume that those genes are candidates genes for positive selection?

Thank you very much for your time, I'm attentive to any comments.

Elisa

Reply all
Reply to author
Forward
0 new messages