Challenges

These tasks are designed to bring together everything you’ve learned this week. Work through them at your own pace and ask for help if you get stuck.

Task 1: Mystery microbe

Make sure your BLAST environment is active before you start:

mamba activate /shared/team/conda/judittalas.mmb-dtp/blast-env

You have been given a 16S rRNA sequence (mystery.fasta) from an unknown bacterium, and a database of 16S sequences from different bacterial species. Your job is to figure out which species it most likely belongs to.

Setup

cd ~/unix_intro/friday_blast
ls

TASKS 1. Have a look at the database file. How many sequences are in it, and which species?

Solution

grep "^>" 16s_db.fasta

You should see six sequences from six different species.

2. Build a BLAST database from 16s_db.fasta.
Hint: these are nucleotide sequences, not proteins.

Solution

makeblastdb -in 16s_db.fasta -dbtype nucl

3. BLAST the mystery sequence against the database and save the results as a TSV file.
Hint: for nucleotide-vs-nucleotide searches, use blastn instead of blastp.

Solution

blastn -query mystery.fasta \
    -db 16s_db.fasta \
    -out mystery_results.tsv \
    -outfmt 6

4. Look at the results. Which species is the best match? How can you tell?

Solution

cat mystery_results.tsv

With only six sequences in the database, the full results are small enough to read directly. The top hit with the highest percent identity is the most likely source species. You should see E. coli at ~99% identity, which is very close but not an exact match. This suggests the mystery sequence is from E. coli, but a different strain to the one in the database.

5. You should notice that most species appear once in the results, but one appears twice. Which one, and why might that be?

Solution

Campylobacter appears on two lines. The other species all align across ~1400 bp, but Campylobacter's best alignment is only ~936 bp, with a second shorter alignment of ~164 bp. The 16S sequence is so divergent between E. coli and Campylobacter that BLAST cannot align the full length in one stretch and splits it into two partial alignments. This is another signal of evolutionary distance on top of the lower percent identity.

6. Looking at the percent identities across all six species, can you group them by relatedness? Which species are more closely related to each other, and which are more distant?

Solution

You should see three tiers:

• E. coli: ~99% identity (closest match, same species)
• Salmonella, Enterobacter, and Klebsiella: ~96-97% cluster. These are all members of the family Enterobacteriaceae.
• Pseudomonas: ~85%. It is a Gammaproteobacterium like the Enterobacteriaceae, but in a different family.
• Campylobacter: ~80%. It belongs to a different class of bacteria entirely (Epsilonproteobacteria), which is also reflected in the shorter alignment length.