Bootstrap analysis stopped (Assertion `tr->likelihood >= initialLH' failed)

75 views
Skip to first unread message

Tommy Harding

unread,
Jan 20, 2016, 9:58:03 AM1/20/16
to raxml
Hi,

I have been running bootstrap analyses with raxmlHPC-PTHREADS-SSE3 version 8.2.4 and often I get this issue where it does a few replicates then stop with the following error:

raxmlHPC-PTHREADS-SSE3: optimizeModel.c:1606: optParamGeneric: Assertion `tr->likelihood >= initialLH' failed.

Aborted (core dumped)


My command lines look like:

raxmlHPC-PTHREADS-SSE3 -T 4 -m PROTGAMMALG4X -n Homologs_m.17881_100BS -s Homologs_m.17881_alignedTrimmedDerep.phy -b 11 -N 100 -k -p 19473


Strangely, when I change the random seed numbers (flags -b and -p) for analyzing the same dataset,  the analysis sometimes gets to completion.
I noticed that this problem was reported before and was apparently fixed in previous RAxML versions. Is it possible that the same issue is popping again in version 8.2.4?


Thanks for your support.

Tommy





Alexandros Stamatakis

unread,
Jan 21, 2016, 4:08:02 AM1/21/16
to ra...@googlegroups.com
can you send me the input data to my private email address please?

I can't promise though that I will be able to deal with RAxML issues
before March unfortunately,

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

Mark Miller

unread,
Oct 17, 2016, 11:23:21 AM10/17/16
to raxml
Hi Alexis, I wondered if this had been addressed?
A user just reported the same issue for v. 8.2.9.

Thanks
Mark

Alexandros Stamatakis

unread,
Oct 17, 2016, 2:33:49 PM10/17/16
to ra...@googlegroups.com
I'll try to check, but since we are in the middle of the from scratch
re-design of RAxML I can't promise that I will look at this soon,

alexis
Reply all
Reply to author
Forward
0 new messages