SH test for topology

[uzm...@fjwu.edu.pk]

Dear all Users,

I need to do SH test on my tree to find out true topology. I found many software/packages doing SH test but didn't found any reliable steps or manual for them which shows which output trees could be used for them like from BEAST and ML best tree could be used. I tried R package but it came out every time with any error mostly. Then I have read raxml manual in which sh-test command is mentioned, but didn't understand: reference tree (-t), trees file (-z) and what would be the output (TEST): it would be a tree file or excel. would you guys help me in understanding this command:

raxmlHPC -f h -t ref -z trees -s alg -m GTRGAMMA -n TEST

I have two trees: one is output of ML (best tree) and another is mcc tree from BEAST analyses. In both, clades arrangement is different and I need to know which give right topology. 

Thank you!

Best,

Uz

[Alexandros Stamatakis]

Dear Uz,

> I need to do SH test on my tree to find out true topology. I found many
> software/packages doing SH test but didn't found any reliable steps or
> manual for them which shows which output trees could be used for them
> like from BEAST and ML best tree could be used.

RAxML needs Newick formatted trees as input.

> I tried R package but it
> came out every time with any error mostly. Then I have read raxml manual
> in which sh-test command is mentioned, but didn't understand: reference
> tree (-t), trees file (-z) and what would be the output (TEST): it would
> be a tree file or excel. would you guys help me in understanding this
> command:
> raxmlHPC -f h -t ref -z trees -s alg -m GTRGAMMA -n TEST

raxml prints out a plain text file containing the test results, it
prints the name of that file in the terminal.

> I have two trees: one is output of ML (best tree) and another is mcc
> tree from BEAST analyses. In both, clades arrangement is different and I
> need to know which give right topology.

The SH test will NEVER tell you which is the right topology, just if the
one with the highest ML score is significantly more likely than the
tree(s) you compare against.

Alexis

> Thank you!
> Best,
> Uz
>
> --
> You received this message because you are subscribed to the Google
> Groups "raxml" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to raxml+un...@googlegroups.com
> <mailto:raxml+un...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.

--
Alexandros (Alexis) Stamatakis

Research Group Leader, Heidelberg Institute for Theoretical Studies
Full Professor, Dept. of Informatics, Karlsruhe Institute of Technology

www.exelixis-lab.org

[uzm...@fjwu.edu.pk]

Thank you Alexis.

Best,

Uz

[uzm...@fjwu.edu.pk]

Hi Alexis,

For SH test, which tree file need to select, either best tree file or bipartite tree file and how could I combine more than two trees for -z. Please help me out.

Thank you,

Best,

Uz

[Alexandros Stamatakis]

Dear Uz,

> For SH test, which tree file need to select, either best tree file or
> bipartite tree file

One or more ML trees.

> and how could I combine more than two trees for -z.

Via the Linux "cat" command.

Alexis

> <https://groups.google.com/d/optout>.

>
> --
> Alexandros (Alexis) Stamatakis
>
> Research Group Leader, Heidelberg Institute for Theoretical Studies
> Full Professor, Dept. of Informatics, Karlsruhe Institute of
> Technology
>

> www.exelixis-lab.org <http://www.exelixis-lab.org>

[uzm...@fjwu.edu.pk]

Dear Alexis,

Thank you. One more question please, how could we get p-value for the test. I got this results:

Tree: 0 Likelihood: -86213.931902 D(LH): -2277.846401 SD: 143.108980 Significantly Worse: Yes (5%), Yes (2%), Yes (1%).

Do we need to infer results by ourselves.

Thank you,

Best,

Uz

[Alexandros Stamatakis]

the p-values are printed here for 5% 2 % and 1%

Significantly Worse: Yes (5%), Yes (2%), Yes (1%).

alexis

> > an email to raxml+un...@googlegroups.com <javascript:>
> > <mailto:raxml+un...@googlegroups.com <javascript:>>.