Hi Sergei,

I'd like to estimate ancestral sequences given a topology, an alignment (AA in my case) and a model (WAG, JTT, REV, maybe LG). Any chance this routine can be isolated from SLAC etc? I thought a bf was there, but I don't see it. The point is to compare with the PAML (aaml) results.

Many thanks,
sergios

Hi Sergios,

There are three different ways to reconstruct ancestors in HyPhy: joint likelihood (default), marginal likelihood and sampled likelihood. If you are happy with the joint likelihood approach, then the BF you are looking for is a post-processor 'Ancestral Sequences': it will become available after you run AnalyzeNucProtData.bf, or any other BF that builds a likelihood function. In the GUI version you can access it via Analysis->Results->Ancestral Sequences; in the command line version use the -p command line argument. The BF itself is in TemplateBatchFiles/post_ancestors.bf.

Cheers,
Sergei

Thanks, Sergei. It worked fine. How could I try the other methods, too, and what is the joint lik reference, please?

Best,
sergios