Should I Fix Branch Length In Paml Analyses?

Question

[qiyunzhu]

Hello friends, I am doing some very regular PAML (codeml) analyses to test positive selection on a gene family. I am confused with the fix_blength option. What should I do with it? I searched around and couldn't get an answer. So far my method is to run M0 model first to get a tree with ML branch lengths, then I use this tree as the user tree and set fix_blength=1 (initial). However, it is very slow comparing to fix_blength=0 (fixed), and sometimes simply fails to finish. I am wondering what is the correct way to set this parameter? Thank you all!

Answer 1

[Lhl]

Have you sorted out?