MrBayes

From MolEvol
Revision as of 14:51, 25 July 2012 by Ejmctavish (talk | contribs) (Datasets and Basic usage)

Getting datasets and instructiions

This tutorial will contain multiple examples of MrBayes uses.

The full tutorial files are available here and the instructions are here
The instructions include telling you how to install MrBayes on your own computer but we will use the cluster version for the lab exercise

Datasets and Basic usage

Log on to the cluster and load up the preinstalled modules with

module load bioware

Download the lab files by using wget

wget https://molevol.mbl.edu/wiki/images/1/10/WoodsHole2012_tutorial.zip

and open the zip file by using

unzip WoodsHole2012_tutorial.zip

or download from here and unzip on your laptop then scp to the cluster

You can get all the instructions needed by using

wget https://molevol.mbl.edu/wiki/images/b/b5/MrBayes_tutorial_instructions.zip 

and unzip in the same way or download from here, unzip and scp to cluster as needed
In the file MrBayes_tutorial.txt we will start the lab from the section ANALYSIS PIPELINE (line 98)
Open the text file on your computer and follow along as you require.

Mr Bayes can be loaded by typing

mb

You can quit MrBayes by typing q and hitting enter

Batch files

A batch file is one that contains a series of commands for a program. Instead of typing commands individually you can run the batch file and it will automatically run all the commands contained within.
This lab consists of individual batch files with MrBayes commands inside. There are 5 sections, each within it's own subfolder of the woodsHole2012_tutorial folder
Inside each of these subfolders is a batch file (conifer_partn.nex, conifer_rjMCMC.nex, conifer_anc.nex, conifer_clock.nex, conifer_dte.nex respectively). This batch file is run in MrBayes by typing

exec batchFilename

within the MrBayes program

  • NOTE: you need to be in the folder that contains the batch file for this to work. Thus, you must cd to the subdirectory before you open MrBayes

Pre-made output

A lot of the analysis we will be performing today is very in-depth and can take a while to run. Sample output has been supplied for you in the output folder within each exercise's directory.
You can try organise it so that, in pairs, you run different exercises so that you can see run times and outputs. You can look through the batch files and the instructions file to see what commands are run

Reading commands in batch files on UNIX command line

You can download the lab onto your laptop and read each batchfile in a plain text editor as you would a normal text file.
If you wish to read the file on the cluster without editing it in nano you can use the following command

more batchFilename

You can use the spacebar to page down through the file and press q to exit. Each line will run on to the next if it hits the edge of the screen so be aware that if commands look strange it could be run over from the line before it
If you are finding the output from more confusing ask a TA about alternative viewers on the cluster.