Tag Archives: BEAST2

Checkpointing Tricks

7 April 2014 by Remco Bouckaert

BEAST 2 stores its internal state occasionally when running an MCMC chain. This means that you can run a chain for a number of samples then resume the chain where it was stopped. This is especially handy when your computer power supply is not very stable or you run BEAST on a cluster that limits the length of jobs.

To resume a run, just select the resume option from the BEAST dialog, or if you run from the command line, use the -resume option

Fast model selection/Skipping Burn-in

Disclaimer: this is for exploratory analysis only. For the full analysis running different chains (say 4) till convergence from different random starting points is the recommended way.

Having said this, exploring different models can take a lot of time, choosing different clock models and their configurations, choosing different tree priors, and substitution models with and without gamma heterogeneity and what have you. Exploring all these settings takes time and often it turns out that many models can be dismissed due to their bad performance. However, every time you change settings you have to go through burn-in. If you set of taxa is small, this is no issue, but if you have many taxa and use models with slow convergence, it may pay to use the partial state of a previous run

For example, we set up a Standard analysis in BEAUti, importing the alignment from examples/nexus/anolis.nex, and change the substitution model to HKY, and leave everything at its default settings, then save the file as anolis.xml. So, we start with a simple strict clock analysis with Yule prior, and HKY substitution model without gamma categories or invariant sites. Run BEAST on anolis.xml for a bit, and it produces the file anolis.xml.state that may look something like this:

birthRate.t:anolis[1 1] (-Infinity,Infinity): 7.55654639037806 
kappa.s:anolis[1 1] (0.0,Infinity): 3.174302336238077 
freqParameter.s:anolis[4 1] (0.0,1.0): 0.27551971155001337 0.26213332703028214 0.11680817105655772 0.34553879036314655 

Now, in BEAUti, you can change the site model to use 4 gamma categories and estimate the shape parameter (click check box with ‘estimate’ next to it), — also, change the log file names — and save as anolis2.xml. If you copy anolis.xml.state to anolis2.xml.state and run BEAST on anolis2.xml it will use the end state for the tree, birth rate, kappa and frequency parameters of the anolis.xml run as the starting state of the anolis2.xml run. For the gamma rate, it will use the setting provided in BEAUti.

You will notice a rise in posterior from about -22428 to -20328 and a rise in log likelihood from about -22453 to -20345 a whopping 2000 log points improvement. The 95% HDP intervals for the likelihood do not even overlap, so adding gamma
heterogeneity helps in fitting the model considerably in this case and the without heterogeneity model does not be considered any further.

The trees are shaped considerably different as well:

Of course, the anolis dataset with only 29 taxa and 1456 characters in the alignment is just a toy example and there is not such a large gain in speed, but for larger analyses re-using a state can help considerably.

Speeding up MCMC (in some cases)

When the tree is sampled from a bad area — which is almost always is initially, if you start with a random tree — the calculation of the treelikelihood can run into numerical trouble. It will switch to ‘scaling partials’, which can result in a severe performance hit. After a while, the tree can move to an area with higher likelihood and scaling is not required any more. One would think that scaling could be switched off after a while, but that is not always desirable. There are a number of schemes, like NONE, ALWAYS, and DYNAMIC, but we are waiting for the scheme that always works. The code promises this scheme:

KICK_ASS("kickAss"),// should be good, probably still to be discovered

but unfortunately, it is commented out.

One thing you can do to speed up BEAST is switch off scaling, but that may result in zero likelihoods initially. Another thing you can do is run the chain initially with scaling, then stop the chain after a little while and resume the chain. The resumed chain will start without scaling and if it is in a good state, it will not need scaling any more and run a lot faster.

What is new in BEAST v2.1.2 and its packages

31 March 2014 by Remco Bouckaert

BEAUti has many bug fixes and enhancements for supporting multiple paritions better. In particular, you can now clone site models and clock models; just select a set of models from the list in the Site Model or Clock Model panel, and the rest of the panel is filled with a “Clone From” selector. This makes it much easier to set all site models to say HKY + 4 gamma categories with estimated frequencies.

The AddOnManager has gotten a facelift. Dependent packages can be automatically installed and uninstalled. You can specify other package repositories than the default one that BEAST uses.

BEAUti can now import FASTA files next to NEXUS files and alignments from BEAST1 and BEAST2 XML files.

SequenceSimulator and LogAnalyser are enhanced to support simulation studies (more on this soon in this blog).

Distributions can be marked stochastic, so when the MCMC inner loop does a sanity check to ensure the incrementally calculated posterior remains the same as a freshly calculated posterior, these distributions can be ignored.

BEAUti automatically updates weights of *BEAST related operators so that the total times spent on *BEAST operators is 20%. This is believed to enhance *BEAST mixing.

New in BEAST-packages

To access packages, you should install them using the install them first.

New in BEAST-labs

BEAST-labs — formerly known as BEASTii — now contains a Script class that allows you to use Javascript (as well as other script languages like python and ruby, provided the appropriate script engine is in the class path). With a Script object you can log complex information that would otherwise require writing a new Java class.

Path sampling/stepping stone analyses is moved to the model selection package.

New in model-selection package

The model-selection package supports path sampling/stepping stone analyses. There is a simple GUI for setting up a path sampling/stepping stone analysis.

The package supports pairwise path sampling/stepping stone analysis where you can estimate the difference of marginal likelihood between two models. This is in general more efficient than estimating marginal likelihoods for each of the two models appart.

New in SNAPP

SNAPP now contains a simple SNP caller — it looks at the first sequence in the alignment, and assumes that the characters there represent the zero value for the site. This means you can use a nucleotide sequence directly in a SNAPP analysis, instead of having to convert it to a binary sequence first.

Tip texts in BEAUti are improved.

New in BDSKY

Robustify against incorrect input, and remove intervalNumber-input.

New packages

Some new packages were added:

On the bleeding edge with BEAST 2: using dev build from command line (and running BEAST faster on windows)

18 March 2014 by Remco Bouckaert

When a bug is discovered that holds you up, and it is relatively easy to fix, the fix may already be in the code but not released yet. The release process takes quite a bit of time, so it is not something we like to do very often. However, you can still use the fixed code (and possibly some new experimental code as well) by using the development build. Every time code is checked in, a new beast.jar file is build and made available for download from here. This file contains all runnable code for running BEAST and BEAUti, but not the BEAUti templates (XML files for setting up analyses in BEAUti). Only if all regression tests pass, the jar becomes available, so that should give some confidence that new code did not break any previous functionality.

I am assuming you already have BEAST installed before, so the Standard and StarBEAST BEAUti templates are on your computer. Note that it is important to store the beast.jar file in the lib folder (details below), so it picks up the BEAUti templates from the correct lcoation. For running BEAST, it does not matter
where you store the jar file.

Using beast.jar on Windows

When you download BEAST you get a zip file. I will assume that BEAST is unzipped on drive C, so there is a directory C:BEAST. If it is located in another place, you will need to replace “C:BEAST” with wherever BEAST is located.

1. Download beast.jar

First, download beast.jar and store it in C:BEASTlib.
The directory C:BEASTlib already contains the beast.jar file that comes with the release, which you might want to back-up. It is not used by BEAST.exe or BEAUti.exe, so it does not affect the installed GUI versions of BEAST and BEAUti.

2. Running BEAUti

To run BEAUti, open the Command Prompt window by clicking the Start button Picture of the Start button, clicking All Programs, clicking Accessories, and then clicking Command Prompt (more here
if you are not familiar with the cmd prompt). In the cmd screen, key in the
following two lines:

cd c:BEAST
java -cp libbeast.jar beast.app.beauti.Beauti

and the familiar BEAUti screen should be shown.

3. Running BEAST

To run BEAST on MyBEASTFile.xml, open a CMD prompt and key in the following:

cd c:BEAST
java -jar libbeast.jar MyBEASTFile.xml

Note: running from the command line tends to make
BEAST run faster on Windows, so that is another reason to use BEAST from the command line.

Note: use java -jar ... instead of java -cp ... as used for running BEAUti.

Using beast.jar on Mac

Assuming you installed BEAST in /Applications/BEAST 2.1.1/ where 2.1.1 is the version number of BEAST. If you installed another version, you have to replace wherever it says 2.1.1 below with that version number.

1. Download beast.jar

First download beast.jar and store it in /Applications/BEAST 2.1.1/lib.

2. Running BEAUti

To run BEAUti, open a terminal and key in the following two lines

cd /Applications/BEAST 2.1.1
java -cp lib/beast.jar beast.app.beauti.Beauti

and the familiar BEAUti screen should be shown.

3. Running BEAST

To run BEAST on MyBEASTFile.xml, open a CMD prompt and key in the following:

cd /Applications/BEAST 2.1.1
java -jar lib/beast.jar MyBEASTFile.xml

Note: use java -jar ... instead of java -cp ... as used for running BEAUti.

Using beast.jar on Linux

Like for Mac, but with /Applications/BEAST 2.1.1/ replaced with the directory where you installed BEAST.