Bootstopping criteria

[Noor White]

Hi all,

I've read through Pattengale et al. 2009 (I use their notation below) and the RAxML v8.2.X manual, but a few of the details of the bootstopping criteria are still unclear. I am specifically looking to run the -#|N autoMRE option (and assuming that the consensus tree criteria in the RAxML manual (page 25) are the same as the Weighted Robinson-Foulds distance-based Criterion (WC) in the Pattengale paper).


-What is the default number of permutations (p) that are run every 50 bootstraps in RAxML (and can p be changed)?

-How are the support values derived in order to calculate the WRF between the bootstrap subsets (c(s1) and c(s2))?

-What does a "threshold for MR-based bootstopping criteria of 0.03" (RAxML manual page 10) mean? Can it be interpreted that the c(s1) and c(s2) trees have no node that differs in support by more than 3%?


Thanks in advance for any clarification on the above.

[Alexandros Stamatakis]

Dear Noor,

> I've read through Pattengale et al. 2009 (I use their notation below)
> and the RAxML v8.2.X manual, but a few of the details of the
> bootstopping criteria are still unclear. I am specifically looking to
> run the -#|N autoMRE option (and assuming that the consensus tree
> criteria in the RAxML manual (page 25) are the same as the Weighted
> Robinson-Foulds distance-based Criterion (WC) in the Pattengale paper).
>
>

> -What is the default number of permutations (/p/) that are run every 50
> bootstraps in RAxML (and can /p/ be changed)?

from the command line help:

--bootstop-perms=number specify the number of permutations to be
conducted for the bootstopping/bootstrap convergence test.
The allowed minimum number is 100!

DEFAULT: 100

In same cases using 1000 is better.

> -How are the support values derived in order to calculate the WRF

> between the bootstrap subsets (/c(s1)/ and /c(s2)/)?

from the consesus supports of the respective tree sets.

> -What does a "threshold for MR-based bootstopping criteria of 0.03"

> (RAxML manual page 10) mean? Can it be interpreted that the /c(s1)/ and
> /c(s2)/ trees have no node that differs in support by more than 3%?

It means that on average the difference is smaller than 3% for 99% of
the permutations.

Alexis

>

>
> Thanks in advance for any clarification on the above.
>
>
>
>
>

> --
> 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

[megje...@gmail.com]

Hi Noor,

I am interested in knowing whether you were successful at specifying the number of permutations with "--bootstop-perms=number".

RAxML simply tells me 'unknown option -- bootstrap-perms...etc'

Any information would be greatly appreciated.

Thanks,

Meg

[Alexandros Stamatakis]

you might be using an old version of raxml, upgrade to v 8.2.9 and it
should work,

alexis

On 30.11.2016 07:21, megje...@gmail.com wrote:
> Hi Noor,
>
> I am interested in knowing whether you were successful at specifying the
> number of permutations with "--bootstop-perms=number".
> RAxML simply tells me 'unknown option -- bootstrap-perms...etc'
> Any information would be greatly appreciated.
>
> Thanks,
> Meg
>
>
>
> On Friday, 29 July 2016 04:11:37 UTC+10, Noor White wrote:
>
> Hi all,
>
> I've read through Pattengale et al. 2009 (I use their notation
> below) and the RAxML v8.2.X manual, but a few of the details of the
> bootstopping criteria are still unclear. I am specifically looking
> to run the -#|N autoMRE option (and assuming that the consensus tree
> criteria in the RAxML manual (page 25) are the same as the Weighted
> Robinson-Foulds distance-based Criterion (WC) in the Pattengale paper).
>
>

> -What is the default number of permutations (/p/) that are run every
> 50 bootstraps in RAxML (and can /p/ be changed)?

>
> -How are the support values derived in order to calculate the WRF

> between the bootstrap subsets (/c(s1)/ and /c(s2)/)?

>
> -What does a "threshold for MR-based bootstopping criteria of 0.03"

> (RAxML manual page 10) mean? Can it be interpreted that the /c(s1)/
> and /c(s2)/ trees have no node that differs in support by more than 3%?

>
>
> Thanks in advance for any clarification on the above.
>
>
>
>
>