LAB B2: rRNA finding, taxonomic classification and multiple sequence alignment

Preparation questions

Study these questions and bring written answers to the lab. See the “Practical information” for more information.

What is rRNA? Which organisms have this?
How does Blast work?
What is “bootstrapping”? How does this procedure estimate the statistical support to the results of an analysis?
What is “worse” in a sequence alignment? Substitutions, small gaps or big gaps?
How does multiple sequence alignment work? Can you explain the “once a gap, always a gap” rule?

Instructions and questions

Welcome to the second computer exercise in bioinformatics! In the previous bioinformatics lab, we learned about the epidemic of unknown bacteria. We have found genes in their recently sequenced genomes, and we saw how to use Blast to compare genes to databases in different ways. Today we will start off again by finding genes, but only of a particular kind: ribosomal RNA (rRNA) genes. Ribosomal rRNA is often used for identifying unknown bacteria at a species-level. Go to the BioToolsBooklet and find a tool for identifying rRNA genes. Run your bacterial genome through it.

Remember to activate the bioinfo environment

conda activate bioinfo

If it does not exist, create it:

conda create -n bioinfo python=3.8

On barrnap installation

If you are getting an error related to “name+” when running barrnap:

run “which barrnap” to find the path of the source code of barrnap
Go to the path, either through the terminal or through the user interface
Open the source code named “barrnap” with a text editor
Search for “name+” with the finder function, and delete the “+”. It should just say “name”.

Questions

Q1 Once you selected your rRNA identification tool from the booklet, apply it to the bacteria files from lab B1 (not B2). How many rRNA genes did you find? What subunit of the rRNA are they (5S, 16S and 23S)? How many of each are there? Hint: Use a flag when running that allows you to save the output file

Open the FASTA results and make a file with just the 16S rRNA genes. This is the subunit that is most commonly used for classification. You will use an online tool for classifying rRNA. Find a tool for doing this in the booklet and run it. Hint: Check the “Search and Classify box”. When the job is complete, press display classification and drag the headers to see the species

Q2 Which matches (species) do you get? Is the classification entirely consistent (do all the matches agree with each other)?

Look up information on this genus online. Wikipedia might be enough, but look up other websites if you’re not convinced.

Q3 Does it make sense that these patients are so sick? Justify your answers and include your sources.

Now that you know a bit more about your bacteria, we can use this information to compare this new isolate to previously sequenced ones. You can download these canonical bacteria in LAB_B2, they will be your database/reference. This might give interesting clues on what is going on with the bacteria in the LAB_B1. Please find in this folder a collection of reference genomes of which one should be your bacterium of interest (from B1).

Q4 Is it likely that very closely related bacterial species will have entire genes added or missing? Why/why not? Hint: consider mechanisms of horizontal gene transfer.

Let’s try to see what distinguishes your new genome from others in the species. One way of finding this (albeit not necessarily the most efficient) is through Blast.

Q5 In which format is the reference genome, nucleotide or protein? What about the file containing the genes you found in the previous bioinformatics lab? In that case, which type of Blast is recommended?

Blast the genes that you found in the previous bioinformatics lab (think about which file to use!) against the reference genome of the bacteria that you chose (reference genomes provided in this lab). Remember to choose a suitable E-value for your search.

Q6 Which genes do NOT find a good match from this database? Retrieve only the non-matching genes and make a separate fasta file with them. **Hint: It’s recommended to use output format 7. See Blast help section for how to choose format.

Depending on the version of Blast, you might get the output in the form of:

# BLASTN 2.9.0+
# Query: xxxxx ID= xxxxx
# Database: xxxxxx
# 0 hits found

To create a file containing only genes that do not have a successful match, you can use grep with the flags B and w. B includes X numbers of lines before the “phrase” in the output file, and w keeps only the lines with the exact “phrase”. Make sure to change X, “phrase” and the filenames to you desired ouput:

grep -B X -w "phrase" result_from_blast.txt > output_file 

Submit the genes WITHOUT good matches to online Blast.

Q7 Which Blast program did you select? Did you change any parameters from the default values? Which and why?

Q8 What are the main Blast hits found? Do they explain the toxicity of this new strain of bacteria?

Q9 From which organism do these genes seem to come from? Does this make sense? Why/why not?

A class of genes that has very important clinical implications is antibiotic-resistance genes. Common mechanisms of antibiotic resistance are pumps that keep the drugs outside the bacterial cell or enzymes that break down the antibiotic, but bacteria can also mutate in a way that makes their own proteins immune to the antibiotics.

You’ve asked for help from the experts at the sequencing center to characterize the antibiotic-resistance genes in your unknown bacteria. They’ve informed you that it is a multi-antibiotic-resistance operon known as mar. It’s mode of action is still unknown, but some things are already understood. The operon contains 4 protein-coding genes, marA, marB, marC and marR. We’re going to work more with these protein products in later labs. For now, let’s take a closer look at the marB genes.

Download the file called MarB.fasta, which contains marB-related proteins from several different bacteria.

We’ll use online Blast again for pairwise sequence comparison. It would take too long to compare each pair of sequences, so we’ll focus on 3 pairs. For each pair you’ll have to choose between blastn, megablast and tblastx as the best tool for aligning them (by best, understand “the tool that gives the most information”). Justify all your answers with your own words as well as the dot-matrix and other pictures from the blast output that you find relevant.

Compare the first sequence, gb|AF226275.1|:1823-2035, with the second, gb|CP003047.1|:1361810-1362022.

Q10 Which Blast tool(s) did you pick? Using this, how closely related are these two proteins?

Now compare the first sequence with gb|CP004887.1|:3422041-3422259.

Q11 Which Blast tool(s) did you pick? Using this, how closely related are these two proteins? How does this answer change comparing different tools?

Finally, compare the first sequence with gi|629665248:741-914.

Q12 Which Blast tool(s) did you pick? Using this, how closely related are these two proteins?

As you’ve noticed, doing pairwise sequence comparisons is quite slow. Fortunately, there are tools for comparing several proteins at the same time. Run this file through a multiple sequence alignment tool. Choose ClustalW format for the multiple alignment, it will make the next steps a little easier.

Q13 Which multiple sequence alignment tool did you choose? Do the results confirm that the sequences in this file are all related? Justify your answer with your own words as well as copying parts of the alignment.

Keep this result output open and run one more multiple sequence alignment tool. Choose the same format as before.

Q14 Which tool did you use now? Do you see any differences in the result comparing the two? Which tool seems better? Justify your answer with your own words as well as copying parts of the alignments.

Now select a relatively conserved part of the multiple sequence alignment and use an online tool for creating a sequence logo.

Q15 What remarkable characteristics do you see in your logo? Can you make any biological hypothesis about this? Include your logo in the answer.

Q16 Which sort of information is highlighted in each of these sequence comparison methods: pairwise alignment, multiple alignment and logo? Can you say in which situation you would pick each of these?

Q17 From a biological perspective, why are there conserved regions or motifs?