BEAST v1.7.5 released

[Andrew Rambaut]

Hi BEAST user,

We pleased to announce the release of minor version update BEAST v1.7.5. Available now from
the download place:

http://code.google.com/p/beast-mcmc/downloads/

The main new feature is the ability to add a starting value for the age of clades specified in BEAUti using
taxon sets. This allows the user to ensure they are within any hard` bounds specified in the priors table. These
starting dates are now used by BEAST when generating an initial tree and will hopefully solve most of the dreaded
"Your initial likelihood is zero" error (which, like most people, I took as a personal insult).

Other things that might be considered new 'features' (although the first sounds more like a bug fix, Marc?):
Conditional reference prior for mutation rates returns a proper density when tree height is
is a random quantity.
Ancestral recombination graph inference improved for two partition problems.
BEAST runs on NSF Keeneland supercomputer in a massively parallel fashion.
BEAUti site model options now has a 'Yang96' button to set a codon position model described in
that paper.
TreeAnnotator now reads MrBayes clock trees.

And a few bug fixes:
Issue 672: BEAUti duplicates rates in phylograms
Issue 671: Exporting taxon names in Nexus format - names with apostrophes are enclosed in single quotes.
Issue 667: Dollo ambiguity code is incorrect when using BEAUti's no data option.
Issue 666 (the issue of the BEAST): BEAUti should read charsets from sets blocks in NEXUS files
Issue 665: The coalescent based simulator doesn't respect hard bounds placed on MRCAs of taxon sets.
Issue 662: Problem with reconstructing synonymous/non-synonymous change counts (BEAGLE must be used).
Issue 659: BEAUti generates various Dollo ambiguity codes

Best,
Andrew

___________________________________________________________________
Andrew Rambaut
Institute of Evolutionary Biology University of Edinburgh
Ashworth Laboratories Edinburgh EH9 3JT
EMAIL - a.ra...@ed.ac.uk TEL - +44 131 6508624


--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

[Andrey Yurchenko]

I have a question. Is there free access to this  NSF Keeneland supercomputer , unfortunately I can't find a information on using BEAST there.  Thank you.

пятница, 15 февраля 2013 г., 22:50:32 UTC+4 пользователь rambaut написал:

[Miller, Mark]

One can access keeneland by requesting an allocation on xsede.

As part of the allocation application you can request time on keeneland. Normally one would request a startup allocation. You would do this at the xsede portal: portal.xsede.org. Startup allocations are awarded within a couple of weeks.

To request an allocation, you must be at a US institution. If you have a US collaborator, they can request the allocation, and put you on the account.

The keeneland machine will give good beast speed ups on certain kinds of data sets, Where the gpu parallelism really helps.

Hope that helps. Feel free to email me directly if you have further questions about applying for xsede.

Best,

Mark

--
You received this message because you are subscribed to the Google Groups "beast-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beast-users...@googlegroups.com.
To post to this group, send email to beast...@googlegroups.com.
Visit this group at http://groups.google.com/group/beast-users?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.