autoMRE bootstraps did not converge

[Noor White]

Hello,

I ran an autoMRE bootstrap analysis on a dataset of 18 taxa, ~900K bp in length, and the bootstraps did not converge after 1000 searches. What is the next step? Is there a flag to run >1000 bootstrap searches?

Thank you in advance for any advice.

[Noor White]

Here is my program call if that is useful:

raxmlHPC-PTHREADS-SSE3 -b 798234 -p 315412 -s dataset.phylip -N autoMRE -m GTRGAMMAI -n bootrep -m GTRGAMMAI -T 40

[Alexandros Stamatakis]

Dear Noor,

Two options:

1. Run more bootstrap replicates (without bootstopping) and apply the a
posteriori bootstrap convergence test that takes as input the set of BS
trees.

2. Investigate why this is not converging, this could be
(i) due to lack of signal or (ii) due to rogue taxa (see
http://sysbio.oxfordjournals.org/content/62/1/162.short).

In fact I had exactly the same issue (bootstrap values not converging)
when writing this book chapter here:

http://www.ncbi.nlm.nih.gov/pubmed/26334924

I quote from that chapter:

"Note that the test dataset used is known to be difficult to analyze. In
particular, it has at least two topologically distinct, yet
statistically indistinguishable, ML peaks (trees) that make its analysis
hard and, as a consequence, yield low support values. Interestingly,
this dataset was not selected on purpose but by chance. Thus, the RAxML
analysis presented here revealed that it is problematic. Tracing back
the origin of this dataset then showed
that the dataset is known to be difficult to analyze (see in-depth
discussion of “data set 1” in Lakner et al., 2008)."

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