Home Pre-workhop Instructions
Lectures
1: Introduction and Key Concepts 2: Variant Detection and Phylogenetic Trees 3: Practical Applications of WGS and Phylogenetics 4: Advanced Applications of WGS
Practical 1: WGS Data Analysis
Obtaining sequencing data Viewing raw sequencing data QC of FASTQ files Cleaning and filtering FASTQs Reference mapping/alignment De novo assembly
Practical 2: Variant Calling and ML Trees
Variant calling and VCFs Building consensus sequences Aligning consensus sequences Maximum Likelihood trees
Practical 3: Timed Phylogenetic Trees
ML timed trees Bayesian trees (BactDating) Bayesian trees (BEAST2)
Practical 4: Transmission and Profiling
Identifying lineages and serotypes Transmission network reconstruction
Practical 5: Mixed Infection, Recombination and ANI
Identifying mixed infection Detecting recombination Average nucleotide identity (ANI)
Practical 6: Phylogeography and Phylodynamics
Phylogeography Inferring population dynamics
Practical 7: Fitness and Selection
Local branching index (LBI) Detecing homoplasy dN/dS GWAS


Maximum likelihood timed trees with TreeTime

These first two exercise will take you through different methods for producing time-calibrated trees using the untimed ML tree produced in the last practical. These methods use the existing topology of the input tree and dates at the tips (sequences) to estimate the date of nodes and re-scale branches per unit time.

First, we will use TreeTime to produce a timed phylogeny by applying a maximum likelihood approach. This is a powerful tool that will not only infer a timed phylogeny but also run an ancestral reconstruction of sequences, estimate past population demographics, and assess the amount of temporal signal in your data by performing a root-to-tip distance regression.

The following files will be used for this exercise:


To run TreeTime we will use the following command:

treetime --tree Taiwan_B117_MLtree.treefile --dates Taiwan_B117_dates.csv --aln Taiwan_B117_aligned_masked.fasta \
	--outdir Taiwan_B117_treetime --clock-rate 0.0008 --clock-std-dev 0.0004 --reroot MN908947.3 --coalescent skyline


Where the parameters are as follows:

Option Description
--tree The input un-timed phylogeny. Here we have used a ML tree but any type (parsimony, neighbour-joining) can be used if branch lengths are in substitution per site.
--dates A CSV file with tip names and dates (format %Y-%m-%d or decimal year) in the first two columns. These can be collection dates or date of symptom onset. Some dates can be partial or missing.
--aln The alignment file used to create the un-timed phylogeny.
--outdir The name to use as the directory to save all output files.
--clock-rate A fixed clock rate to use in substitutions/site/time. This option can be removed if the clock rate can be confidently estimated from the data.
--clock-std-dev The standard deviation on the clock rate in substitutions/site/time. Only to be used if using the --clock-rate parameter.
--reroot Re-root the tree to the named sequence. This can also be set to 'oldest' to re-root on the sequence with the oldest date.
--coalescent Activates the Kingman Coalescent model. This can be set as constant or skyline, or left blank to run without a tree prior.

There are other parameters that can be included when running TreeTime, for full details please see here.


Explore the output from TreeTime, there will be 11 files in the "Taiwan_B117_treetime" folder:

  1. Open "ancestral_sequences.fasta":

    2. Discuss what you think these sequences are, why are there more sequences (rows) than tips in the tree?

  2. The tree files are saved in the two .nexus files, as well as a plotted tree in PDF format. In FigTree, or using any other tree viewing software, open the two .nexus files:

    3. What is the difference between these trees?

  3. Finally, open the "skyline.pdf" file:

    4. What does this plot represent?

More information on the effective population size can be found here


Next activity: Two-step Bayesian timed trees with BactDating