Can we say a gene is under positve selcetion if the LRT tests results is significant but BEB detects no positively selective sites?

[Duo Xie]

Hi Prof.Yang and everyone,

When I use the optimized branch-site model to detect positively selected genes, if a gene is significant in the LRT test, but I found no positively selected sites in mlc file using BEB method, can we say this gene is under positive selection?

Thank you!

Duo

[Ziheng]

that may be a sensible conclusion.  one possibility is that collectively there is evidence for positive selection for the whole gene, but the information or evidence for each individual site is too weak.

you can look at the rst file which has the posterior probabilities for all sites, to see whether that is the case.  the mlc file lists only those with posterior higher than 0.5.

ziheng