Hi HyPhyers, I am using AnalyzeCodonData to examine dN dS patterns in several genes in a clade of fishes.  In the output, we have the list of results like this:

givenTree.Node82.nonSynRate=0
givenTree.SCARUS_SPINUS.synRate=9.51535
givenTree.Node82.synRate=4.43235
givenTree.Node56.synRate=3.27021
givenTree.Node56.nonSynRate=0
givenTree.SCARUS_SPINUS.nonSynRate=0

And then we have the E[Syn subs/nucleotide site] tree and the E[Non-syn subs/nucleotide site] tree:  and the dS and dN trees.  

Can you briefly tell me the differences between, for example SynRate at a node, Syn subs/nucleotide site branch lengths, and dS branch lengths?

I am also interested in how one would back calculate the actual number of S or N substitutions on a particular branch, and which would be best to use for dN/dS ratio.  

Finally, I cannot use the menu item dN/dS for branches no matter how complex the model I use- any idea why not?

Thank you, HyPhy is an excellent package.

Dear mwestneat,

You should be able to find the answers to most of your questions on pp 12-14 in Multimedia File Viewing and Clickable Links are available for Registered Members only!!  You need to Login.

The direct estimate for the dN/dS ratio is the nonSynRate/synRate ratio for a given branch.

I am not sure what you mean by the actual number of S, N substitutions - one way to estimate that would be to reconstruct ancestral states using ML and count substitutions using the most parsimonious path. Is that what you had in mind?

SLAC (implemented in HyPhy and also described in the link above) will count these numbers on a per-site basis.

Cheers,
Sergei