(c) 2016 Justin Bois. This work is licensed under a Creative Commons Attribution License CC-BY 4.0. All code contained herein is licensed under an MIT license.
This exercise was generated from a Jupyter notebook. You can download the notebook here.
Exercises 2.3 and 2.4 were inspired by Libeskind-Hadas and Bush, Computing for Biologists, Cambridge University Press, 2014.
For all problems except for the git practice, we will work with real data from the Salmonella enterica genome. The section of the genome we will work with is in the file ~git/bootcamp/data/salmonella_spi1_region.fna. I cut it out of the full genome. It contains Salmonella pathogenicity island I (SPI1), which is contains genes for surface receptors for host-pathogen interactions.
If you have not already, finish the practice exercises with git.
There are packages, like Biopython and scikit-bio for processing files you encounter in bioinformatics. In this problem, though, we will work on our file I/O skills.
a) Use command line tools to investigate the FASTA file. You will notice that the first line begins with a >, signifying that the line contains information about the sequence. The remainder of the lines are the sequence itself.
b) Use the file I/O skill you have learned to read in the sequence and store it as a single string with no gaps.
Pathogenicity islands are often marked by different GC content than the rest of the genome. We will try to locate the pathogenicity island(s) in our section of the Salmonella genome by computing GC content.
a) Write a function that divides a sequence into blocks and computes the GC content for each block, returning a tuple. The function signature should look like
gc_blocks(seq, block_size)
To be clear, if seq = 'ATGACTACGT' and block_size = 4, the blocks to be considered are
ATGA
CTAC
and the function should return (0.25, 0.5). Note that the blocks are non-overlapping and that we don't bother with the end of the sequence that does not fit completely in a block.
b) Write a function that takes as input a sequence, block size, and a threshold GC content, and returns the original sequence where every base in a block with GC content above threshold is capitalized and every base in a block below the threshold is lowercase. You would call the function like this:
mapped_seq = gc_map(seq, block_size, gc_thresh)
For example,
gc_map('ATGACTACGT', 4, 0.4)
returns 'atgaCTAC'. Note that bases not included in GC blocks are truncated.
c) Use the gc_map() function to generate a GC content map for the Salmonella sequence with block_size = 1000 and gc_thresh = 0.45. Where do you think the pathogenicity island is?
d) Write the GC-mapped sequence (with upper and lower characters) to a new FASTA file. Use the same description line (which began with a > in the original FASTA file) and have line breaks every 60 characters in the sequence.
a) Write a function, longest_orf(), that takes a DNA sequence as input and finds the longest open reading frame (ORF) in the sequence (we will not consider reverse complements). A sequence fragment constitutes an ORF if the following are true.
ATG.TGA, TAG, or TAA.Note that the sequence ATG may appear in the middle of an ORF. So, for example,
GGATGATGATGTAAAAC
has two ORFs, ATGATGATGTAA and ATGATGTAA. You would return the first one, since it is longer of these two.
Hint 1: The statement for this problem is a bit ambiguous as it is written. What other specification might you need for this function?
Hint 2: You might need to check the documentation of different string methods, like find() to see how to get the functionality you are looking for.
b) Use your function to find the longest ORF from the section of the Salmonella genome we are investigating.
c) Write a function that converts a DNA sequence into a protein sequence. The dictionaries in the bioinfo_dicts module may be useful.
d) Translate the longest ORF you generated in part (b) and perform a BLAST search. Search for the protein sequence (a blastp query). What gene is it?
e) [challenge] Modify your function to return the n longest ORFs. Compute the five longest ORFs for the Salmonella genome section we are working with. Perform BLAST searches on them. What are they?