Hello,
The dn/ds ratios of external vs internal branches of a genealogy are compared to detect transient deleterious polymorphism in viruses (in particular because it could inflate the estimated rate of substitution). This is generally implemented in PAML. However, it is also possible to compare the dn/ds ratios of external vs internal branches with HYPHY as described in the chapter of the Phylogenetic Handbook (Lemey et, al, 2009) dedicated to selection pressure (Kosakovsky Pond et al, page 475): "Use TestBranchDNDS.bf from the Positive Selection rubrick of Standard analyses". Alternatively, can GAbranch from DataMonkey be used with the same objective: compare the dn/ds ratios of external vs internal branches? Are these three methods relying on the same evolutionary models and - even if they are not - are they likely to give similar results? With many thanks for your help. Denis.

PS. I can run the TestBranchDNDS.fr with HYPHY1 (win32 on Windows) but not with HYPHY2.

Hi Denis,

For this type of analysis you should probably also look at IFEL (also described in the book you site) and implemented on Datamonkey. The short answer to your question -- the models used by PAML and HYPHY tend to be conceptually similar (codon substitution models), but many details (e.g. how they are parameterized, whether or not dS can vary from site to site) are quite different. Generally speaking, you should expect some differences in results from the two packages.

What happens when you try to run TestBranchDNDS.bf in HYPHY v2?

Sergei

Hi Sergei,

Many thanks for your informative (and prompt) answer.

Regarding TestBranchDNDS.fr with HYPHY2, the following error message appears:
"« Could not reach batch file C:\Documents and Settings\fargetd\Bureau\HYPHY2.0
mp\HYPHY2\TemplateBatchFiles\TemplateModels\chooseGenetic.Code.def
Path stack:{Documents and Settings\fargetd\Bureau\HYPHY2.0\tmp\HYPHY2.0\,C:\Documents and Settings\fargetd\Current Task has been terminated. Would you like to see the remaining error messages, if there any?"

This error message appears even if I set HYPHY2 at the root of C, whether I use my own data or the example datasets, if I load the data or tree files before the analysis, if I put HYPHY2, the batch file and the data file together within the same file...

I suspect that the same problem happens with other standard analyses. Then I wondered whether I was doing something wrong, but I do not have any problem running the same analysis with HYPHY1(win32). I do not have any problem either using "New analyses" and the graphical interface with HYPHY2.

Please let me know if you need further details. With many thanks for your help. Best Regards. Denis.

Hi Denis,

This sounds like a path problem (which I would like to fix).
What is that path name (e.g. C:Documents and Settings\fargetd\Bureau\HYPHY2.0\) of where the HyPhy distribution is?
It is important NOT to move the executable itself from the path where it was installed (otherwise it won't find the needed files).

Sergei

Hi Sergei,

Please find below the path.
C:\Documents and Settings\fargetd\Bureau\HYPHY2.0\tmp\HYPHY2.0

I loaded and extracted HYPHY2 on my desktop (Bureau means Desktop). I did not move the executable from the file where it was installed.

I loaded HYPHY2 three times to make sure that this was not a problem due to internet communications: the same problem arises.

Please let me know if you need further details. With many thanks for your help. Best Regards. Denis.

I couldn't help but notice this in your error message:
..."TemplateBatchFiles\TemplateModels\chooseGenetic.Code.def"

Is this a typo?

- a.

Hi Denis,

Looks like an old bug (treating "\t" as "Tab" in Windows path names has resurfaced). Try renaming the 'tmp' folder where HyPhy is as something that does not begin with the letter 't' and the error messages should disappear. I'll fix the issue for the next code update. Thanks for letting us know about this error.

Sergei

Hi Denis,

I fixed the bug -- please download the HYPHY v2 binary for Windows and confirm that the code works now.

Thanks,
Sergei

Hi Serguei,

I have just downloaded the latest HYPHY v2 binary for Windows. It is working fine now. Thank you very much for your help. Denis.