A collection of potential questions for the final, submitted
by students in this class and the previous offering of this class.
Variants of at least three of these will be on the final examination.
Last time, a few students noted that Thinking of potential exam
questions is helping me study.
Broad Questions
What is bioinformatics?
What is an algorithm?
Examine one bioinformatics technique we have studied this semester
and give specific suggestions for improvements that could refine the
technique.
Papers
What are the primary characteristics that distinguish a strong paper
in Bioinformatics from a weaker paper? Give examples to support your
argument.
Compare and contrast the initial T. Rex paper and the Neanderthal one
with respect to findings and methods.
Kellis et al. discuss the comparative analysis of closely related
genomes to find genes and their regulatory mechanisms. List two other
gene finding techniques that were in use before the publication. Explain,
in brief, the rationale and methodology in finding a gene with specific
insight into ORFs, motifs and gene conservation. Why did they choose
different methods for finding genes as opposed the ones they used for
finding their regulatory networks? List three results from the paper
and explain their significance.
After hearing Baggerly's talk, some members of a certain political party have
increased their distrust of science. They say It looks like scientists
not only fake data, but also make trivial mistakes in their analyses that
lead to spurious conclusions.
They have hired a consulting firm to
identify potential areas of fraud and trivial mistakes
in the
Kellis et al. paper. What areas might that consulting firm point
out as likely locations?
Microarrays
Explain a bioinformatics problem that can be solved using
Microarrays. Explain a situation where Microarray data can be inaccurate.
Sequencing and Sequences
Briefly describe Sanger sequencing, Direct sequencing, and Shotgun sequencing.
Why is sequence assembly so difficult?
Draw the steps of shotgun sequencing.
Given a PAM matrix, some short sequences, and a simple BLAST-like
algorithm, explain and demonstrate the process by which the algorithm
would unite the sequences into a single longer sequence. Why is this
process important?
Gene Prediction
Describe three methods for gene prediction in eukaryotes.
How do you identify the genes in a sequence? Give three different
algorithms and explain.
Sequence Comparison
What are the reasons we would want to compare sequences?
Why would we want to compare protein sequences instead of DNA sequences?
How does the Needleman-Wunsch tool differ from BLAST? Give algorithmic and functional reasons.
Phylogeny
How are ORFs used to study evolutionary relatedness? What computational
tools are required for this process?
Explain why the different programs/algorithms we used to determine
phylogenetic trees came up with different trees for the same batch of
species given.
Final Projects
As part of your project, you developed some Python code to aid in the
analysis of biological data. Present and explain that code.
Explain your final project as you understand it. What was something that
didn't go according to your plans or expectations during the course of
your project, and how did you adapt to that problem?
Summarize each of the final projects.
Chou-Fasman
Given a sequence of amino acids and a portion of the Chou-Fasman table,
identify any alpha-helices, beta-strands, or beta-turns which the
Chou-Fasman algorithm would find.
Describe the Chou-Fasman algorithm, specifically the implementation of
the alpha helix and beta sheet discovery.
Long Questions
Give a hypothesis and some data. However, make the data not fit the
hypothesis, or worse yet, not have a clear interpretation whatsoever. Or
we could attempt to align two sequences that have no good alignment. This
would be really mean, but also a good way to prepare us for the real
lesson that we've been through over and over again: this is a messy,
messy subject, and we will have some problems and beliefs that just won't
lead us anywhere. Of course, some of us toyed with this on our projects,
but I think it would be interesting. You could also ask us to agree or
disagree with the given hypothesis (citing reasons) and develop a new
one if necessary.
You were able isolate a bacteria species, R. rudolphi, from the north pole.
a. During microscopic studies you noticed that in the presence of fog
rudolphi cells begin turn red and aggregate. How would you investigate
the cellular changes happening in the presence of fog? What prior
information is necessary? What limitations will your analysis face?
b. You identify the individual gene responsible for the redening of
rudolphi cells. What biological mechanisms could be controlling
the expression rate of the red genes?
c. Your colleague believes the gene was inherited by a horizontal
gene transfer event. Given its complete genome, how would you test
this hypothesis? You can describe programs you would use, programs
you would write, or any other method you might apply.
You are on a board that allocates grant money to bioinformatic
research. After an exhausting series of allocation decisions the board
has enough cash to fund one of two proposed projects. The first research
team promises to increase the accuracy of of protein fold prediction
algorithms. The other research team promises to complete a technique
that would dramatically reduce the time and cost of genome sequencing,
making it available to the general public. Which project do you choose
to fund? Justify your choice, making sure to acknowledge the arguments
of those who would make the opposite choice.
You have just finished sequencing the first completed genome of a
newly discovered bacteria species isolated from agricultural soil.
a. What methods would you use to determine the phylogeny of the
bacteria using its complete genome sequence? Are there any advantages
or disadvantages?
b. You now want to understand the content of the genome. How would
you identify and classify genes? What measures would you take to ensure
your interpretations are accurate?
c. Your genomic analysis reveal a novel gene coding for an unknown
protein. How would you analyze the protein's gene sequence to determine
its properties? What are the strengths and weaknesses of your methods?
Miscellaneous
Summarize the projects of our two guest speakers.
Define horizontal vs. vertical gene transfer. What are the similarities
and differences between them?
Why are viruses dangerous? Why is it not easy to make a vaccination that
will prevent a virus from multiplying?
What are the differences between PAM and BLOSUM?
What is the difference between dominant vs. recessive genes?
The prediction of protein structure seems like a difficult problem to
solve as there are many multidisciplinary teams all attempting to tackle
the problem. Name several biological techniques that have arisen. Then
name some computational approaches that have been used to resolve this
problem. Analyze their strengths and weaknesses.
;
