LAB B1: Gene finding, Blast and sequence alignment

The scenario

There is an outbreak of mysterious infectious diseases in your town. Doctors do not know what is causing it, but they do know it is spreading fast. Patients come into the hospital presenting very different symptoms and doctors suspect that there might be more than one pathogen behind it. There are three variations of the mysterious disease:

Disease 1 mostly affects children and their families. It is a severe form of diarrhoea that can lead to death if the patient isn’t carefully rehydrated. Many, but not all, patients also present vomiting.
Disease 2 seems to be sexually transmitted. It causes the skin or mucosa of the affected areas to form blisters, similar to burning. These open wounds in the anal-genital area often lead to other infections. Many patients have to take intravenous serum instead of normal food, and some must be kept under total isolation in the intensive care unit.
Disease 3 has been spreading the fastest. Patients initially have flu-like symptoms, but in a matter of days or weeks it evolves to a life-threatening pneumonia, associated to chest pain and bloody coughs.

Doctors have been able to isolate bacteria from the blood of a few patients from each of these conditions (healthy individuals usually do not have any bacteria in the blood stream). They have extracted DNA from the bacteria and had it sequenced. DNA sequencing produces very small reads that have to be assembled (put together) into longer fragments of contiguous bases, called contigs. A bioinformatician at the sequencing centre has already done the initial work. It is now up to you and your colleagues to find out as much as possible about this pathogen. The patients are counting on you!

Preparation questions

We will start each lab discussing a few preparatory questions. You will not gain or lose points from them, but you might be called to discuss them in front of your classmates, so be prepared!

What is genome annotation? What is its goal?
What is an Open Reading Frame? How does this relate to genome annotation?
What does it mean to “align two sequences”? What is the goal?
What is a p-value? How is that different from an e-value (expectation value) in BLAST?

Instructions and questions

Select one of the files bacteria1.fasta, bacteria2.fasta, or bacteria3.fasta, which correspond to diseases 1, 2 and 3, respectively. Make a folder called bioinformatics in your local computer account, download the fasta file and save it there.

Q1 Which of the 3 unknown bacteria have you chosen to work with?

Open the fasta file in a text editor such as gedit. Alternatively, view it through the command line.

Q2 How is a fasta file organised? What information can be found in it? Is this a practical format? Why/why not?

To understand the metabolism and life-cycle of an unknown species based on its DNA content, we have to study the functions of its genes. The first step for doing so is finding the gene sequences within the genome. Fortunately, there are tools that can find the genes inside a genome, based on certain sequence characteristics. Check the Bioinformatics Tools Booklet and look for tools for gene finding.

Q3 Which tools did you find?

Take a look at their websites. Feel free to explore them for a few minutes. Then, pick one tool to use in this assignment. It’s important to have the nucleotide sequences as output.

Q4 Which tool did you choose? Why? Did you change any parameters from the default settings? Which, how, and why?

Now you have a fasta file of all the candidate genes you’ve found.

Q5 How many genes did this tool find? Is this a good estimate for the number of genes in this organism? Why/why not? Hint: check the Bioinformatics Tools Booklet for commands that can count lines containing a pattern

Now that we have identified the genes (at least the most likely genes according to the gene finder you employed), we can start studying their functions. One approach for doing this is to compare these new sequences with annotated sequences from other organisms. A very popular tool for doing this is Blast. Check the Bioinformatics Tools Booklet for instructions on how to use online Blast for nucleotides. Open the fasta file with the candidate genes and Blast the first 5 genes.

Q6 Which Blast variant have you chosen? Why? Did you change any parameters from the default settings? Which, how, and why?

Q7 How many hits did you find for each gene? What do they correspond to? Does this make sense? Note: if there are too many hits, just describe the top ones!

Q8 Considering how many genes you have found, is it practical to examine the function of each corresponding protein by online Blast?

Blast can also be run locally, through the command line interface. This allows whole genomes or collections of genomes to be scanned very fast. However, it takes a bit more bioinformatics expertise to go through the very large files that are produced, so we’ll do this in a slightly simplified way this time. Let’s look for RNA-polymerases, that is, the enzymes that transcribe DNA into RNA. They have already been downloaded from NCBI as described here, using “Bacteria” and “RNA polymerase” as keywords. You can retrieve this file directly as polymerases.fasta Now look into this fasta file. A lot of the sequences are described as “CDS”.

Q9 What does that mean? What is the difference between CDS, EST and ORF?

Q10 How many sequences are there in the fasta file? This can be quickly counted through the command line

The first step to running Blast through the command line is to prepare a database. You have the necessary fasta files to compare, ie, your unknown bacterium and the collection of bacterial RNA-polymerases. One of these files is going to be your database, and the other one contains all of your queries (the sequences to be identified).

Q11 Which of these files should be the database, and which one should be the query? Why? What would happen if you did it the other way around?

Look into the Bioinformatics Tools Booklet to see how to prepare a Blast database.

Q12 Which command did you run? Describe what each part of it does.

Now that the database is ready, it’s time to run nucleotide Blast.

Q13 Which command did you run? Describe what each part of it does.

Q14 From all the results you got, pick 3 matches and give its e-value and bit-score.

Now download the file fewer_polymerases.fasta. This file contains only 50 of the sequences from the polymerase database. Format a Blast database from this file, too, and run Blast using it.

Q15 Which commands did you run? Are there any differences compared to what you did before? Did you get the same hits as you had before?

Now look at the e-values and compare it to what you had before.

Q16 Is there any change between the e-value you had before and what you got now? How do you expect the e-value to change with the size of the database? Hint your result may or may not follow what is expected

Not all polymerases are the same. Let’s compare a few of the ones in the database. You can find all of them in the smaller dataset. For this, you can go back to using online Blast.

For the following sequences, justify your answer with your own words, but also include a dot plot and at least part of the sequence alignment. Compare the first two sequences in the file, gi|328835344 and gi|328835342

Q17 Which type of Blast did you run? Why?

Q18 How similar are these sequences? Looking at the sequence headers, is this expected?

Now compare the first sequence, gi|328835344, with the one that has ID gi|619834969.

Q19 How similar are these sequences? Looking at the sequence headers, is this expected?

Q20 Considering what you learned from statistics, but also your biological knowledge, is this match relevant? Why/why not?