These exercises are based on real
data, and simulate situations that are likely to arise for an experimental
biologist working in the miRNA discovery field. The exercises are open ended
and there is more than one way to answer the questions.
Exercise 1: Characterization of a potential miR
From a human sarcoma cell line you
have cloned the following sequence, which you suspect could be a miRNA:
>RNA1taggtagtttcatgttgttgg
Q1.1: Where is the sequence encoded in the
genome? Is it found in more than one place?
(Hint: The BLAT
search at UCSC is useful for mapping sequences to the genome.
NB! The aim is to find a perfect match, not to search for homologs)
Q1.2: Is the sequence conserved in other
organisms?
(Hint: Conservation track in the UCSC browser)
Q1.3: Could it be a miRNA? If so, is it
new, or is it already known?
(Hint: extract surrounding sequence and fold it with RNAfold)
To study the function of RNA1 you
decide to inhibit it with antisense LNA. After
transfecting the anti-RNA1 LNA into your cell culture you perform an Affymetrix
microarray experiment to look for changes in global gene expression. You
observe that many genes are differentially expressed after transfection,
and that the most strikingly upregulated gene in the
experiment with the inhibitor has this Affy_probe_set_id:
221278_at.
Q1.4: What is
the corresponding gene and the mRNA sequence for this Affy_id?
(Hint: You can use the UCSC Genome Browser)
Q1.5: Has this gene been
associated with microRNAs before?
(Hint, search the literature, e.g. in PubMed)
Q1.6: Why do we see increased
signal from probe 221278_at when RNA1 is inhibited?
Q1.7: What targets is the miRNA
predicted to have? Is 221278_at among them?
Q1.8: Could there be a link
between the miRNA and suffering from Trichotillomania, obsessive-compulsive pulling your hair?
(Hint: look in OMIM for the target gene)
Q1.9: What consequences could you
envisage, if the miRNA was knocked out in vivo?
Exercise 2: Target prediction and evaluation
You wish to
identify the targets of the miRNA: hsa-let-7e
Q2.1: How many targets
do you identify by using three popular prediction methods?:
miRANDA
Targetscan
Pictar
(Hint: there are
links to each method from miRBASE)
Q2.2: How well do
the predictions agree? Draw a Venn diagram showing the overlap!
(Hint: First, the
gene lists should be made comparable, e.g. by DAVID. Next, you may generate the
Venn diagram by e,g, using
the “Venn diagram
creator”.
Q2.3: What would you suggest doing to validate
the predicted targets?
Exercise 3: 454 discovery
of novel miRs – or not?
The following
five sequences were identified from a 454-run on breast cancer samples:
ACTCGGCGTGGCGTCGGTCG
ACTCGGCGTGGCGTCGGTCGT
ACTCGGCGTGGCGTCGGTCGTG
ACTCGGCGTGGCGTCGGTCGTGG
ACTCGGCGTGGCGTCGGTCGTGGT
Q3.1: Do these
sequences represent novel miRNAs?
Q3.2: If so, how
many?
Q3.3: How can the
data be interpreted in terms of miRNA biogenesis and processing?