Skip to content

Tutorial 1: BEAUti

Jump to bottom
Walter edited this page Nov 18, 2022 · 4 revisions

Codon Substitution Models Tutorial

Walter Xie (18 November, 2022)

Introduction

This tutorial introduces the BEAST 2 CodonSubstModels package for codon substitution models through a simple tutorial. The tutorial involves understanding of models describing evolution at the codon level allowing the estimation of measures of the selective forces acting on proteins (Kosiol et. al. 2000).

You will need the following software at your disposal:

  • BEAST, BEAUti, TreeAnnotator and other utility programs. This tutorial is written for BEAST v2.7.x, which is available for download from beast2.org.

  • Tracer - this program is used to explore the output of BEAST (and other Bayesian MCMC programs). It graphically and quantitively summarizes the distributions of continuous parameters and provides diagnostic information. At the time of writing, the current version is v1.7.

  • FigTree - this is an application for displaying and printing molecular phylogenies, in particular those obtained using BEAST. At the time of writing, the current version is v1.4.3.

You also need a basic knowledge of using BEAST and BEAUti. The tutorials are available from beast2.org/tutorials or Taming the BEAST website.

Installation

The CodonSubstModels package can be installed from BEAUti's package manager (File => Manage Packages), using the Install/Upgrade button. Please remember to restart the BEAUti after you install the CodonSubstModels package.

Package Manager

BEAUti

The CodonSubstModels template has to be used (File => Template => CodonSubstModels) before you are going to setup an analysis from the BEAUti. Make sure the title of BEAUti on the top is changed to "CodonSubstModels".

Load Template

Loading the NEXUS file

Load the example NEXUS file codon.nex (click to download) from File => Codon Alignment.

Partitions

After you choose the file, there will be a pop-up menu to ask the genetic code. This dataset contains nucleotide sequences of 2 humans, 2 chimpanzees and 2 gorillas, so we need to choose Vertebrate Mitochondrial.

Genetic Code

All codes are defined by the standard genetic code tables from GenBank. Please make sure you are selecting the correct genetic code, otherwise some sequences may incorrectly translate into stop codons (*). As a consequence, it causes the exception in BEAST.

If the data is valid, you will see the partition appeared in the panel. This example dataset has 6 taxa and 333 triplets. The data type is codon.

Partitions

Setting the codon substitution model

Now choose the Site Model panel to specify codon substitution model for your analysis. At the moment, only M0 is available, which is also known as GY94, first published by Goldman & Yang 1994, and late designated to be one of the M-series (M0-M13) proposed by Yang et al. 2000. We keep both omega and kappa the same as the default setting.

The type of equilibrium codon frequencies πj consists of "equal", "F1X4", "F3X4", and "F60/F61" (Yang 2006). We choose F60/F61 which is also the default value for the codon frequencies model.

Codon Site Model

Setting the clock model

The rest of models are using the same models in the standard BEAST analysis including the clock model. In this analysis we are going to leave the selection at the default value of a strict molecular clock.

Priors

Then, we simply change the tree prior to Coalescent Constant Population, and keep the other priors as the default.

Priors

Setting the MCMC options

In this exercise, setting 500,000 for Chain Length and 1,000 for Log Every will be good enough to provide reasonable ESS. You can also rename the log to m0.log and tree log to m0.trees.

MCMC

Save the xml in the end.

Next -> BEAST

Clone this wiki locally