Hi Darrell

to process multiple files you can use a wrapper file.

Code:

_rrOptions = {};

SetDialogPrompt ( "Provide a list of files to process:" );
fscanf          (PROMPT_FOR_FILE, "Lines", _filesToProcess );

SetDialogPrompt ( "Provide a list of outgroups:" );
fscanf 		(PROMPT_FOR_FILE, "Lines", _outgroups );

_rrOptions ["00"] = "Nucleotide/Protein";

_rrOptions ["03"] = "HKY85";
_rrOptions ["04"] = "Global";

for ( _fileLine = 0; _fileLine < Columns ( _filesToProcess ); _fileLine = _fileLine + 1 ) {
	_rrOptions ["01"] = _filesToProcess [ _fileLine ];
	_rrOptions ["02"] = _outgroups [ -fileLine ];
	ExecuteAFile ("/Applications/hyphy/HYPHY/TemplateBatchFiles/RelativeRate.bf", _rrOptions );
}

 

this file will require two input files: one with the exact paths to the files to process, and another with the outgroups for each of the input files. Obviously these need to be in the same order. To get the first you could use the unix find command.

Code:

 find `pwd` -name "*.nex" > nexus.files  

If the outgroup is the same for all the files you could remove the second fscanf in the wrapper and simply hardcode the name of the outgroup with
Code:

 _rrOptions ["02"] = "myoutgroup";  

and remove the corresponding line from the for loop.

You will also need to edit the location of the batch file in the ExecuteAFile command.

I think that answers your third question.

re the others

1. you need to set the ti/tv parameter to global.

in the case of HKY85 we use

Code:

global R=0.3250146192987597;HKY85={4,4};
HKY85[0][1]:=R*a;
HKY85[0][2]:=a;
HKY85[0][3]:=R*a;
HKY85[1][0]:=R*a;
HKY85[1][2]:=R*a;
HKY85[1][3]:=a;
HKY85[2][0]:=a;
HKY85[2][1]:=R*a;
HKY85[2][3]:=R*a;
HKY85[3][0]:=R*a;
HKY85[3][1]:=a;
HKY85[3][2]:=R*a;

 

here R represents the ti:tv and is shared across all branches. a is a branch length parameter independent for each branch.

2. I'm not sure we have a syn and non-syn relative rates implementation. if you look at the result from ExportLikelihoodFunction there are constraints placed on the branches.

Code:

constrained3TaxaTree.Ingroup2.a=0.05138303096204919;
constrained3TaxaTree.OutGroup.a=0.3247689253900513;
constrained3TaxaTree.Ingroup1.a:=constrained3TaxaTree.Ingroup2.a;
 

in order to do syn and non-syn relative rates tests, you would need to
i. set up a codon model
ii. have separate branch length parameters for syn and nonsyn rates
iii. place constraints on syn and nonsyn rate independently

something like

Code:

constrained3TaxaTree.Ingroup2.synrate=0.05138303096204919;
constrained3TaxaTree.OutGroup.synrate=0.3247689253900513;
constrained3TaxaTree.Ingroup1.synrate:=constrained3TaxaTree.Ingroup2.synrate;

 

It wouldn't be too difficult to implement, but may be duplication if it is already done. Sergei is on leave till Monday. Let me bounce it off him when he returns.

hope this helps somewhat

cheers ./w

Thank you for responding so quickly, but I still have a problem.  Each of my files has three sets of nucleotide data.  The first two are ingroups, and the last is the outgroup.  Is there any simple way to tell it to use the third set in each file as the outgroup instead of looking in a different file?  The outgroup name will always begin with ">ENSMUST00".  Also, where can I look at the documentation for the options for RelativeRate.bf?  Thanks again for your help.

Alternatively if you can assign user-defined internal node names in your tree files outside of HyPhy, and provide that info via some sort of lookup table then you're in business.
And the HBL command you'll need is RerootTree(treeID,branchID).
-A.

Yay, it works, and by using:
_rrOptions ["00"] = "Codon";
_rrOptions ["01"] = "Universal";
_rrOptions ["04"] = "GY94";
_rrOptions ["05"] = "Local";
_rrOptions ["06"] = "nonSynRate";

LIKELIHOOD_FUNCTION_OUTPUT = 1;

I can get my non-synonymous rates too.  You guys should post a documentation about making wrapper batch files on the Hyphy website, along with a few examples, its much easier than editing the existing files. 

Thanks for all the help