optimal number of initial trees

[Victor Lima]

Hi everyone !

I am trying to make a phylogeny with 3200 pb and 6300 terminal taxa. I wonder if there would be appropiate to consider just 20 initial trees under new rapid hill-climbing algorithm (default) or change either number of initial trees or algorithm?

I will appreciate advisement.

Cheers,  

[Alexandros Stamatakis]

Dear Victor,

I'd run 20 ML searches initially with the standard algorithm and then
calculate the RF distances as well as some statistical tests (SH etc.)
to understand what the tree space looks like.

For such datasets you should probably run more inferences because the
tree space is very rough and then summarize the ML trees that are not
statistically significantly different from each other using consensus
tree methods.

Cheers,

Alexis

> --
> 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
Adjunct Professor, Dept. of Ecology and Evolutionary Biology, University
of Arizona at Tucson

www.exelixis-lab.org

[Victor Jiménez]

Thanks Alexis,

I thought get the Best ML tree under default parameters (20 Parsimony trees ) and then make 100 bootstrap calculation. So I used the CIPRES portal but 24 hours were not appropiate and I got just 43 bootstrap trees. I know I could not be the best. 

Please tell me if I am right, when I calculate the Best ML tree, the program makes rearrangements under some determined algorithm (hill-climbing / Parsimony by default) and selects just one of these as the BEST. 

Are you telling me that I have to repeat this process 20 times to obtain 20 BEST ML Trees and then compare them with SH and RF? Or change the default algorithm to obtain just one BEST ML?

I will thank your advisement.

Many thanks,

--
You received this message because you are subscribed to a topic in the Google Groups "raxml" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/raxml/FuJNQ8XMz0I/unsubscribe.
To unsubscribe from this group and all its topics, send an email to raxml+un...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

Blgo. Victor Jiménez Vásquez 
Facultad de Ciencias Biológicas 
Laboratorio de Sistemática Molecular y Filogeografía
Universidad Nacional Mayor de San Marcos

Lima, Perú

[Alexandros Stamatakis]

Dear Victor,

On 18.07.2015 00:15, Victor Jiménez wrote:
> Thanks Alexis,
>
> I thought get the Best ML tree under default parameters (20 Parsimony
> trees ) and then make 100 bootstrap calculation. So I used the CIPRES
> portal but 24 hours were not appropiate and I got just 43 bootstrap
> trees. I know I could not be the best.
>
> Please tell me if I am right, when I calculate the Best ML tree, the
> program makes rearrangements under some determined algorithm
> (hill-climbing / Parsimony by default) and selects just one of these as
> the BEST.

Yes, but it also prints out the other trees.

> Are you telling me that I have to repeat this process 20 times to obtain
> 20 BEST ML Trees and then compare them with SH and RF? Or change the
> default algorithm to obtain just one BEST ML?

No, please have a look at the RAxML manual, there is an option that
allows you to just do as many ML searches as you want to without doing
bootstraps, that is also available on CIPRES I think.

Alexis

>
> I will thank your advisement.
>
> Many thanks,
>
> 2015-07-17 15:12 GMT-05:00 Alexandros Stamatakis

> <alexandros...@gmail.com <mailto:alexandros...@gmail.com>>:

> <mailto:raxml%2Bunsu...@googlegroups.com>
> <mailto:raxml+un...@googlegroups.com
> <mailto:raxml%2Bunsu...@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
> Adjunct Professor, Dept. of Ecology and Evolutionary Biology, University
> of Arizona at Tucson
>

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

>
> --
> You received this message because you are subscribed to a topic in
> the Google Groups "raxml" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/raxml/FuJNQ8XMz0I/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> raxml+un...@googlegroups.com

> <mailto:raxml%2Bunsu...@googlegroups.com>.

>
> For more options, visit https://groups.google.com/d/optout.
>
>
>
>
> --

> Blgo. Victor Jiménez Vásquez
> Facultad de Ciencias Biológicas
> Laboratorio de Sistemática Molecular y Filogeografía
> Universidad Nacional Mayor de San Marcos
> Lima, Perú
>