HyPhy message board - Print Page

HyPhy message board
http://www.hyphy.org/cgi-bin/hyphy_forums/YaBB.pl HYPHY Package >> Feature Additions >> AU test http://www.hyphy.org/cgi-bin/hyphy_forums/YaBB.pl?num=1227998487 Message started by hannah on Nov 29^th, 2008 at 2:41pm

Title: AU test
Post by hannah on Nov 29^th, 2008 at 2:41pm

Hi all,

Perhaps I missed it, but is there an implementation of the approximately unbiased test for comparing tree topologies? It would be great to have the equivalent of CONSEL, showing the p-values from various tests of competing topologies.

Thanks!
Hannah

Title: Re: AU test
Post by Sergei on Dec 1^st, 2008 at 7:55am

Dear Hannah,

HyPhy only implements the Shimodaira Hasegawa (or its simpler version for 2 trees - Kishino Hasegawa) test. Which AU test are you referring to? Can you provide a reference?

Cheers,
Sergei

Title: Re: AU test
Post by hannah on Dec 1^st, 2008 at 2:01pm

The test is described in:

An Approximately Unbiased Test of Phylogenetic Tree Selection
Author(s): Hidetoshi Shimodaira
Source: Systematic Biology, Vol. 51, No. 3 (Jun., 2002), pp. 492-508

I am not at all an expert... but I've seen this test used in publications. I'd love to hear any thoughts you have on this vs. SH (e.g. for processing GARD results).

Thanks!
hannah

Title: Re: AU test
Post by Sergei on Dec 2^nd, 2008 at 8:15am

Dear Hannah,

I recall reading this paper years ago (it is very good), but I never got around to implementing the test in HyPhy. Sorry:(

Cheers,
Sergei