Dear Sergei,

As per your suggestion I tried to run the compare_selection script in HyPhy. I realized that it expects the two alignments to be of the same length. I want to know, is it not possible to compare selection pressures between two paralogues or orthologues that differ in sequence length?

Secondly, can't we compare hyphy's FEL, REL or the codeml's site, branch and branch-site model estimates of omega for different paralogues or even orthologues directly, making sure that we are comparing the estimates of the same models? If we find a drastic accumulation of non-synonymous mutations relative to synonymous ones in all these models that are based on different methodologies, then isn't that indicating a greater influence of positive selection in the particular gene in comparison with its homologue?

Thank you,
David