Background: For many diseases with known causative mutations, screening methods have been developed to detect whether people have a high risk of becoming sick, even before the onset of the actual disease.

Over the last few years, the cost of full genome sequencing has gone down so that, in some cases, it might be cheaper to collect the complete genome sequence of patients with a high risk of carrying variants associated with the disease, rather than using targeted screening procedures.

Cystic fibrosis is a complex disease, where patients often manifest the following symptoms: problems with lung functions, diabetes and infertility. From a genetic point of view, there are several mutations associated with this disease. The gene CFTR (short for Cystic Fibrosis Transmembrane Conductance Regulator) encodes an ion channel protein acting in epithelial cells and is encoded on chromosome 7 of the human genome. CFTR carries several non-synonymous genetic variants, some of which leading to premature stop codons that are known to cause the disease.

---

Goal: In this assignment, you have access to the human reference genome (chromosome 7) as well as the full genome annotation. In addition, you have genome sequence data of chromosome 7 from five individuals from a family at risk of carrying mutations related to the disease.

Your task is to write a Python program that will extract a disease-causing transcript from the CFTR gene, translate the gene sequence to its corresponding amino-acid sequence and based on the reference amino-acid sequence determine whether any of the five given individuals is affected.

Download the lecture slides from here.

Download the appropriate files

The main task is divided into several steps. The first step is to download the reference sequence file and the appropriate reference annotation file from the Ensembl database:

Human reference DNA for chromosome 7 (fasta format):

• Homo_sapiens.GRCh38.dna_sm.chromosome.7.fa.gz

Human reference annotation file (GTF format):

• Homo_sapiens.GRCh38.93.gtf.gz

If you are not familiar with the file formats, read up online on how the files are structured. For example, here you can find a short description of the different (tab-delimited) fields of a GTF file.

Some of the tasks involve outputting long sequences. To make sure they are correct, use the utils.check_answers package (from the downloads folder from the course topics website). You can import it that way:

from utils import check_answers

More detailed instructions are given with each task that uses the package.

Warmup

1. Make a directory for the project for you to work in and put the project files there.

   Tip

   The commands below are for Mac and Linux and should also work on Windows Subsystem for Linux. You can get help in the terminal by writing the command name followed by --help, such as cd --help. Naturally you can also search the web!

   1. Download the files (see instructions above)
   2. Open a terminal navigate to a directory where you want your project to be. Use cd to change directory and pwd to print the working (current) directory.
   3. Make a folder for the project. Use mkdir or a file explorer. Check that the files are there using by listing the directory's contents with ls.
   4. Move the files to this project. Move files using mv. Again, use ls to see that the files end up where you want them to be.
   5. Unpack any compressed files: the file extension .gz indicates gzip compression which can be decompressed using gunzip. The command for .zip files is unzip.
   6. Examine the file contents using cat, head and tail.

2. What is the length of chromosome 7 on the reference sequence?

   Tip

   Open the reference fasta file and read it line by line. Study the example in the lecture!

   In a loop, ignore the first line and get the length of each following line.

   Don't forget to remove the trailing newline character from each line.

   Sum up all the lengths you found.

   Answer

   Chromosome 7 has 159,345,973 base pairs.

3. How many genes are annotated in the GTF file?

   Tip

   You need to understand the structure of a GTF (gene transfer format) file for this project. Take your time and read up on the file format if you are not sure how to solve this task.

   To get the number of genes, open the GTF file and read it line by line.

   In a loop, count all features of type gene.

   Answer

   There are 58,395 genes annotated in the GTF file.

Architect a method

All following tasks are related to the CFTR gene.

In the annotation file (the GTF file), the CFTR gene has the id ENSG00000001626 on chromosome 7.

1. How many transcripts can the CFTR gene generate?

   Tip

   Again, think about the structure of the GTF file.

   Open the GTF file.

   In a loop, count all transcript features for the gene.

   Answer

   The CFTR gene can produce 11 different transcripts.

2. Which of these transcripts is the longest transcript in nucleotides?

   Tip

   Open the GTF file.

   Fetch the start and stop positions for each transcript of the CFTR gene to calculate its length.

   Keep in mind that sequence numbering starts at 1 in the GTF file format.

   Answer

   The transcript with the id ENST00000003084 is the longest of 11 transcripts and spans 188,703 bases.

3. Fetch the DNA sequence for that transcript.

   Tip

   Open the reference fasta file.

   In a loop, ignore the first line and append each line to a list, removing the trailing newline character.

   Outside the loop, use the join function to concatenate the lines from the list to get the reference sequence. Avoid concatenation inside the loop, as it is slow and wastes memory.

   Extract the start and stop positions of the longest transcript to fetch its DNA sequence from the reference sequence, but think about where the index starts from.

   Answer

   Write your results to a file and compare it to the correct result using check_answers.ex3("resultsFile.txt").

   The entire sequence can also be found here.

4. Fetch the DNA sequences of all exons for that transcript, spliced together to one sequence.

   Tip

   First, you need to save the start and stop positions of all exons of the longest transcript.

   Then you can use a similar loop to the one you used in task 3 to extract the DNA sequence of each exon.

   Finally, you need to concatenate the DNA sequences of all exons.

   Answer

   Write your results to a file and compare it to the correct result using check_answers.ex4("resultsFile.txt").

   The correct sequence can also be found here.

5. What are the start positions and sequences of the start_codon and stop_codon for that transcript?

   Tip

   Find the start_codon and stop_codon features of the longest transcript in the GTF file, including the start positions of the start- and stop-codon.

   Check that the start_codon has the sequence ATG, and that the stop_codon corresponds to a proper stop codon.

   Make your program print a warning message in case the transcript does not begin with a start-codon and end with a stop-codon.

   Answer

   The start codon has the sequence ATG and starts at position 117,480,095.

   The stop codon has the sequence TAG and starts at position 117,667,106.

6. Translate the above sequence of all exons into amino acids, using an implementation of the translation table from the utils.rna package (from the downloads folder from the course topics website).

   Tip

   Translate the DNA sequence of the concatenated exons into amino acids from the start codon position of the transcript on.

   Answer

   Write your results to a file and compare it to the correct result using check_answers.ex6("resultsFile.txt")

   The correct sequence can also be found here.

Find the patients at risk

We are reaching the goal for this assignment!

A mutation in the transcript ENST00000003084 causes a premature stop codon to be introduced into the amino acid sequence. This creates a truncated protein, causing cystic fibrosis. Use the python code you have written to solve the tasks above and extend it to compare the reference genome sequence of chromosome 7 to the sequences of the following 5 patients in fasta format (Patient1.fa.gz, Patient2.fa.gz, Patient3.fa.gz, Patient4.fa.gz, and Patient5.fa.gz) to determine which of them are carrying a mutation in the CFTR gene that causes a truncated protein.

Extra task

1. Use BioPython to parse the fasta file and to translate DNA nucleotides into amino acids.

   Tip

   Check the BioPython tutorial on how to parse a fasta file with BioPython.

   Read up on the built-in translation method and the BioPython translation tables.

Solution

Here are some possible solutions to the assignment. There are of course many correct solutions, we only present one of the alternatives.

1. Dan’s solution from the last day
2. Notebook solution
3. Standalone script