Phylogeny Exercise¶

Author: Vi Varga

Course: Applied Bioinformatics (BBT045)

Due Date: 23.02.2023

Introduction¶

In this exercise you will practice your newly acquired skills in phylogenetic inference and "tree thinking", by analyzing the evolutionary history of a gene family. Please note that unlike the tutorial, this exercise should be completed on your own. If you have questions, please reach out to me via e-mail (virag.varga@chalmers.se) or message me on Canvas. The assignment should be uploaded to Canvas as a Jupyter Notebook (.ipynb file) by 23.59 on Friday, February 23, 2023, along with the requested supplementary files.

If you do not already have this Jupyter Notebook downloaded (i.e., if you haven't cloned the GitHub repo, here), you can use wget to download the Jupyter Notebook version of this file from here.

In this exercise you are going to follow the phylogenetic analysis workflow discussed during the lecture and tutorial, starting with collecting the data necessary for running the analysis (in the form of homologous protein sequences from different species), through the interpretation of the results (i.e., comparing species trees and gene trees).

For this exercise, you will be working with species within a phylum of protists (single-celled eukaryotes) known as the Apicomplexans. Apicomplexans are mostly obligate parasites, meaning that they rely entirely on their host organism to survive. There is debate over the possibility of some species from this phylum being host-associated but non-parasitic, and some recently-discovered Apicomplexans are thought to be free-living, though they are not well-studied (del Campo et al. 2019).

Phylogeny of Apicomplexa

A Newick tree for the clades of interest is provided below (as well as in the Apicomplexa_clades.nwk file):

When visualized as a cladogram, the tree looks like this:

Phylogenetic tree of key clades in Apicomplexa

Below are listed the species you will be working with:

Porospora gigantea
Plasmodium falciparum
Babesia microti
Toxoplasma gondii
Cryptospodidium parvum

Which clade of Apicomplexa do each of the above species belong to? List them below.

You might recognize some of these names! Phylum Apicomplexa contains a number of medically important parasites:

Cryptosporidium species cause Cryptosporidiosis, a water-borne diarrheal disease that often affects immunocompromised people.
Babesia species cause Babesiosis, a tick-borne illness that causes flu-like symptoms. Since the parasite infects and destroys red blood cells, it can also cause anemia.
Toxoplasma gondii causes the disease Toxoplasmosis, which humans contract from food and/or being around pets that are infected. Most infections are asymptomatic, though it can cause flu-like symptoms and serious complications for the fetus if contracted during pregnancy, up to stillbirth or miscarriage.
Plasmodium falciparum is a causative agent of Malaria, a disease which still kills more than half a million people each year. Among the Plasmodium species, P. falciparum causes the most deadly strain of the disease.

The Gregarines are comparatively mild parasites - they infect invertebrates, and are rarely fatal to the organisms they infect.

Why might it be important to study organisms like the Gregarines? (Hint: Take another look at the phylogenetic tree of Apicomplexa.)

Before moving on to the actual phylogenetic workflow, you should also decide what the outgroup will be. This won't actually come into play until the visualization stage, but it's always good practice to be thinking ahead when following a research pipeline.

Which group of organisms should be the outgroup? Why?

You may be wondering, "Why not select an outgroup from elsewhere in Alveolata (the superphylum that Apicompla belongs to)?" After all, the Gregarines are considered part of Apicomplexa - aren't they too closely related to the rest to be useful in a study like this?

In most cases, you might be right! A group of organisms like those in Chromerida, which are also in Alveolata but not considered Apicomplexans, might be a better choice!

Phylogeny of Alveolata, showing Chromerida and Apicomplexa

In this case, we'll be using the Gregarines because of the gene families presented to you as options - all of these gene families are involved in the "glideosome," which is a key component of Apicomplexan pathogenicity. The glidesome is involved in parasite attachment to, movement along and invasion of host cells. More distantly related species within Alveolata may not have copies of these genes. For example, C. velia doesn't have a copy of the ROM4 gene (Boisard et al. 2022).

Obtaining & Exploring Data¶

For this exercise, you will be obtaining your own data from the NCBI database.

You will start by selecting a gene family to study. Below I've listed a few options:

ROM4
AKMT
MyosinA
DGK1

Since Plasmodium falciparum and Toxoplasma gondii have really well-annoted genomes, search for the gene sequence of your choice on the NCBI with, for example: "Plasmodium falciparum[orgn] ROM4". Save the header and amino acid sequence to a FASTA file on your local machine. I've provided recommended search combinations below to make things go a bit more smoothly; the members of these protein families do not all include these gene family names in the actual names of the individual genes themselves on the NCBI. If there are multiple results, just go with the first gene hit.

Search recommendations:

ROM4: P. falciparum ROM4
MyosinA: P. falciparum Myosin-A
AKMT: T. gondii apical complex lysine methyltransferase
DGK1: T. gondii diacylglycerol kinase accessory domain

Save the FASTA sequence of your selected reference protein representative of your chosen protein family to a file on your local computer. This file should be included in your assignment submission.

Next, you'll need to obtain homologous gene sequences from the NCBI database. BLAST your protein sequence on the web to accomplish this, and choose the "FASTA (complete sequence)" download option. This file should be included in your assignment submission.

Note that you can select specific organisms to BLAST against on the NCBI - this significantly reduces the amount of time a search takes, and in our case, makes sure you get only the information you're looking for. There are of course many members of the Apicomplexa that we will not be working with here - no need to sift through all of the gene sequences from those organisms. Include only our 5 species of interest; note that you'll need to select both the A and B versions of P. gigantea individually.

Open your multiple sequence FASTA and describe some of its characteristics. How many sequences are in the file? How many representatives of each species are there (count the two P. gigantea species as only one)?

Describe the following examples with all words that apply from among: homologous, orthologous, paralogous

a. Two copies of MyosinA in B. microti

b. Copies of AKMT in P. falcigarum and T. gondii

Before proceeding on to the construction of the MSA, clean your sequences.

Do you notice any differences in the files? What does this tell you?

Multiple Sequence Alignment¶

Now that you've familiarized yourself with the data, it's time to create a multiple sequence alignment.

Write the code you used to create the MSA below. Explain the meanings of the flags, and why you chose to use them.

In [ ]:

#Fill in your code here!

This MSA file should be included in your assignment submission.

Now open the MSA file in your alignment viewer of choice.

What patterns do you observe?

Creating the Tree¶

Now it's time to create a phylogenetic tree. Since you're less pressed for time than you were in-class, use the model finder option of IQ-TREE, if you wish. However, if you'd prefer to use the same tree model as you did during the tutorial, that's fine, too!

As you may remember from the in-class tutorial, it has turned out to be rather difficult to run IQ-TREE from within Jupyter Lab, as the program will only be able to access 1 core, even if you've booked two. Therefore, I will instead ask you to run a batch job.

While we have not discussed them in much detail, batch jobs are what we call the jobs submitted to the backend of an HPC (in this case, Vera) using the SLURM job scheduling system. In order to run IQ-TREE as a batch job, please follow these steps:

Copy the batch script template named sbatch_iqtree.sh from the /cephyr/NOBACKUP/groups/bbt045_2024/Phylogeny/Homework/ directory back to your working directory.
Modify the sbatch_iqtree.sh file that you have copied to your working directory, by filling in the necessary information:
- Your Chalmers email (line 7)
- The full path to your working directory (i.e., where this Jupyter Notebook and your homework files are) (line 34)
- The name of your cleaned MSA file (line 41)
- The prefix/file base name you want the IQ-TREE output files to have (line 61)
Submit the batch script with sbatch sbatch_iqtree.sh in your regular Vera terminal (i.e., not from within a terminal you open in Jupyter Lab)
Make a record of the job ID that is printed in your terminal when you submit your job. This job ID is necessary in order to be able to trace back any errors/issues, or cancel a running job.

In the cell below, fill in the job ID that was printed out when you submitted the job:

In [ ]:

%%script false --no-raise-error
sbatch sbatch_iqtree.sh
# FILL IN WHAT TEXT IS RETURNED ON THE COMMAND LINE HERE

Which tree model did IQ-TREE pick for you, if you used the model finder? Why is the model finding process important?

The PREFIX.treefile and PREFIX.log files should be included in your assignment submission.

Now open the PREFIX.treefile in a phylogenetic tree visualizer of your choice.

What does it look like to you? Do you notice anything?

Edit the alignment one time, either manually or using a tool. Save the new MSA, and run IQ-TREE on it again, using the model that IQ-TREE previouslsy determined to be the best fit for you. The new MSA file, PREFIX.treefile and PREFIX.log files should be included in your assignment submission.

What method of alignment editing did you choose to use? Does your alignment look different in an alignment viewer? Does the resulting tree look different? How so? If you used a command-line tool, include the code below:

In [ ]:

#Fill in your code here!

Comparing the gene tree with the species tree¶

Finally, compare the two gene trees you have visualized with the species tree given to you. (The R script provided to you during the tutorial may help with this, but feel free to do it your own way!) Include .png/.svg/.jpeg/etc. versions of both of your visualized phylogenetic trees in your assignment submission.

How do the gene trees compare to the species tree? Do you notice any interesting patterns?

Is this a good gene family for creating a species tree? Why or why not?

Citations¶