Class presentation
For your class presentation, you need to pick a paper,
read it, present
it to the class (with a PDF that will be posted on the class
webpage), and answer questions. The presentation should
take 15-20 minutes, and will be followed by up
to 15 minutes for questions.
Your grade on the class presentation will be content
and your ability to answer questions (that would be based
only on understanding the paper), so don't worry about
how pretty your slides are.
You can pick a paper you like, but I will need to approve
the choice (I will make sure the paper is within the scope
of the course and not too difficult for you to be able to present it).
By March 15,
send me a list of 2-4 papers you would like to
present, along with the
hardcopies of the papers; I will select one of these papers.
Also, I want everyone to pick a different paper, so if you have
requested a paper that someone else has already been approved for,
I'll not be able to approve that assignment for you. Therefore,
the earlier you send me your requests, the better!
You will need to send me the PDF for your presentation at least
one week before the presentation; I will send your presentation
back to you with comments, which you can incorporate if you wish.
The final presentation (in PDF form) is due 48 hours before
your presentation, so I can upload it.
Probably the easiest thing to present is a method
addressing something we've studied (or will study). Here is a list of
interesting methods that might be good for class presentations.
- ML tree estimation methods: IQtree, RAxML,
PhyML, FastTree -2
- Bayesian tree estimation methods (e.g., BEAST,
MrBayes, etc.)
- Distance-based tree estimation methods: FastME, NINJA,
FastTree (not the same as FastTree-2)
- Multiple sequence alignment methods: Clustal-Omega,
Contralign,
Muscle, MAFFT, Opal, SATCHMO, PROMALS, PROMALS3D,
Prank, Probcons, Probalign, T-COFFEE
- Alignment and tree co-estimation methods:
BAli-Phy, Alifritz, and an unnamed method
introduced in a PNAS paper by Bouchard-Coté and Jordan
- Multi-locus phylogeny estimation:
BEST,
*BEAST,
GLASS,
iGLASS,
iGTP,
MP-EST,
MulRF,
PhylDAWG,
SMRT-ML, STAR, STEAC, SVDquartets
- Alignment-free phylogeny esimation methods
- Genome rearrangement methods (even just computing
the rearrangement distance between two genomes)
- Whole genome alignment (pairwise and multiple are both
interesting)
- Phylogenetic networks
- Some genome assembler (e.g., Euler)
- Metagenome taxon identification and abundance
profiling (e.g., Carma, Kraken, Megan, Metaphlan, MetaPhyler,
NBC,
PhyloPythia, Phymm, and PhymmBL)
- Software to correct a gene tree with respect to a species tree
(e.g., Notung by Dannie Durand and
an unnamed method
by Rasmussen and Kellis)
- Software/methods to detect orthology
- Methods to date ancestral nodes in a tree
- Methods to infer ancestral sequences in a tree
Another kind of paper that would be fun to present would be
one that addresses some controversial topic (which abound
in this area). Some of those topics are:
- Concatenation vs. coalescent-based methods.
See, for example, the recent paper in PLoS Currents: Tree of Life
(Tonini et al., 2015) and
papers by John Gatesy and Mark Springer (most
recently in Molecular Phylogenetics and Evolution)
- Concatenation vs. supertree methods
- Restricting data to "good" loci, or using
all the loci
(see, for example, Rokas et al. Nature 2003,
volume 425, number 6960, pages 784-804, and
Salichos and Rokas Nature 2013, volume 497, number
7449, pages 327-331)
- The role of morphology vs. molecular sequence
data
- Can trees be reconstructed in the
presence of horizontal gene transfer? (See
Gogarten et al. Molecular Biology
and Evolution Volume 19, pages 2226-2238, 2002, and
discussion and other cited papers in Roch and Snir,
Journal of Computational Biology 20(2):93-112, 2013.)
Another type of paper that would be of interest is
one that addresses the use of these methods to answer
some biological problem. If this is of interest to you,
please talk to me about what you would want to present.