Difference between revisions of "BEAST"

Revision as of 23:37, 25 January 2021

Description

BEAST is a cross-platform program for Bayesian MCMC analysis of molecular sequences. It is entirely orientated towards rooted, time-measured phylogenies inferred using strict or relaxed molecular clock models. It can be used as a method of reconstructing phylogenies but is also a framework for testing evolutionary hypotheses without conditioning on a single tree topology. BEAST uses MCMC to average over tree space, so that each tree is weighted proportional to its posterior probability. We include a simple to use user-interface program for setting up standard analyses and a suit of programs for analysing the results.

Required Modules

modules documentation

Serial

beast

System Variables

HPC_BEAST_DIR - installation directory

Galaxy

Beast is also available for use through the friendly web interface of the UFRC Galaxy instance. It's located in the "Phylogenetics" section.

Interactive beauti use

In short, you have to use a SLURM GUI session and specify the module you want as the default beast module could be different from the one you'd like to run.

E.g.

module load gui
gui start --module beast/2.6.0 -e beauti

See GUI Programs for reference.

Some versions of BEAST/BEAUTI may be available in OnDemand.

Additional Information

Use -beagle_GPU beast command-line argument to invoke BEAST with GPU acceleration when running in a GPU Job
To use CPU acceleration the beagle library use the '-beagle_SSE -beagle_instances N' command line switches where 'N' equals to the number of requested Beagle library instances.
To limit the number of computing cores Beast will use for posterior computation to the number requested from the batch system use the '-threads N' where 'N' is the number of threads. Then, the job resource request should be equal to nodes=1:ppn=X where X = N + 1.
General performance considerations - http://www.beast2.org/beagle-beast-2-in-cluster/.
Beast2 plugins are installed into your personal directory. This means that you need to start Beauti and install plugins manually.

@@ Line 26: / Line 26: @@
 ==System Variables==
 * HPC_{{uc:{{#var:app}}}}_DIR - installation directory
-* HPC_BEAST_DOC - docs directory
-* BEAST - Beast directory
-==GPU Accelerated BEAST==
-The latest BEAST-1.x installs have a 'beast-gpu' wrapper that can greatly accelerate some analysis when run in the <code>gpu</code> partition on HiPerGator. 7-20x speedups were observed on a medium-sized test dataset allowing to shorten cycle time for that particular analysis from 1 hour per million states to under 15 minutes and in some cases to under 3 minutes per million states for the majority of the run. Look for the 'beast-gpu' command in a particular beast module like beast/1.10.0.
 ==Galaxy==
@@ Line 39: / Line 35: @@
 E.g.
   module load gui
-  gui star --module beast/2.6.0 -e beauti
+  gui start --module beast/2.6.0 -e beauti
 See [[GUI Programs]] for reference.
+Some versions of BEAST/BEAUTI may be available in [https://ood.rc.ufl.edu OnDemand].
 <!--Additional-->
 {{#if: {{#var: exe}}|==Additional Information==
-* To use the beagle library use the <code>'-beagle_SSE -beagle_instances N'</code> command line switches where 'N' equals to the number of requested Beagle library instances.
+* Use -beagle_GPU beast command-line argument to invoke BEAST with GPU acceleration when running in a [[GPU_Access|GPU Job]]
+* To use CPU acceleration the beagle library use the <code>'-beagle_SSE -beagle_instances N'</code> command line switches where 'N' equals to the number of requested Beagle library instances.
 * To limit the number of computing cores Beast will use for posterior computation to the number requested from the batch system use the <code>'-threads N'</code> where 'N' is the number of threads. Then, the job resource request should be equal to <code>nodes=1:ppn=X</code> where X = N + 1.
 * General performance considerations - [http://www.beast2.org/beagle-beast-2-in-cluster/ http://www.beast2.org/beagle-beast-2-in-cluster/].