Home

This site has material for a software workshop on phylogenetic networks, used at the 2018 & 2019 MBL workshop on molecular evolution (earlier version for a 2016 workshop).

It covers steps to go from a bunch of multiple alignments (aligned gene sequences, or loci) to a phylogenetic network displaying the relationships between the species in the alignments

topics covered

• example data to download
• TICR pipeline overview: from sequences to quartet concordance factors (CFs, proportion of genes having a particular history)
• PhyloNetworks for network estimation: from quartet CFs or gene trees to phylogenetic networks, and bootstrap analysis to understand network uncertainty
• TICR test: is a population tree with ILS sufficient (vs network)?

Set-up

locally

• Download BUCKy
• Download TICR
• Download QuartetMaxCut
• Download MrBayes
• Download julia and follow instructions to install julia
• Install the necessary packages: open julia then type

  using Pkg # to use functions that manage packages
  Pkg.add("PhyloNetworks") # to download & install package PhyloNetworks
  Pkg.add("PhyloPlots")
  Pkg.add("RCall")      # packaage to call R from within julia
  Pkg.add("CSV")        # to read from / write to text files, e.g. csv files
  Pkg.add("DataFrames") # to create & manipulate data frames
  Pkg.add("StatsModels")# for regression formulas
  using PhyloNetworks   # may take some time: pre-compiles functions in that package
  using PhyloPlots

  and close julia with exit().

for MBL workshop

Login to your particular virtual machine (VM) using the IP address on the sticker attached to the back of your name tag. If, for example, your IP address was 123.456.789.321, you would type the following into the terminal on your local computer (i.e. your laptop) and press the enter key:

ssh moleuser@123.456.789.321

After login, you want to copy the phylo-networks in your home directory:

cp -r moledata/phylo-networks ./
cd phylo-networks

You can open julia by typing julia in the terminal.

more details

• TICR pipeline

  • analyze each locus with MrBayes
  • do a concordance analysis with BUCKy on each set of 4 taxa
  • summarize all quartet concordance factors (CFs)
  • estimate a species tree using Quartet MaxCut

  We will not cover an alternative pipeline (which you could use outside the workshop) to

  • analyze each locus with RAxML, including bootstrap
  • estimate a species tree with ASTRAL

• Julia package PhyloNetworks with SNaQ (Species Network applying Quartets)

  • read the quartet concordance factors
  • estimate a phylogenetic network
  • choose the number of reticulations
  • run a bootstrap analysis
  • plot bootstrap support for tree edges, and for hybridizations

• TICR test in the R package phylolm (Testing Incongruence Checking in R)

  • test if a population tree with the coalescent (ILS only) is adequate to explain the quartet concordance factors
  • identify taxa involved in outlier quartets