Hi, Sergei,

I'm working with dNdSRateAnalysis.bf to perform REL on many alignments.

1. I'd like to perform model comparisons to test the relative strengths of selection on the different alignments, almost identical to the procedure you outline on pp. 28-29 of your book chapter. I see that I can select a non-positive discrete distribution to test for positive selection. Is there an option for inputting other constraints, such as specific averages for the discrete rate classes or proportions of sites in them?

2. I'm scripting all of this. After I select "custom" (option 2) to run a subset of tests, is there a syntax I can use that will allow me to run more than one of the tests (but not all)?

Thanks!

Sarah

Dear Sarah,

1). dNdSRateAnalysis.bf (as implemented) does not permit you to specify parameter values a priori. You would have to change the source code a bit (e.g. read in values from a file) following Model definition

2). Yes. If you look towards the end of the book chapter (Multimedia File Viewing and Clickable Links are available for Registered Members only!!  You need to Login) for how to script existing analyses with ExecuteAFile ("file", options), you can craft the options list to deal with multiple selections. You do this by listing all the selections in order and ending with an empty string, e.g.


Code (]
options["05"):

= "Option 1"; options["06"] = "Option 5"; options["07"] = "";

HTH,
Sergei