Assignments for Wednesday

 

• Work through Olga's example of Bayesian thinking (here)

• Explore the population genetic simulations at http://www.radford.edu/~rsheehy/Gen_flash/popgen/.
• Using the same fitness and frequency for the A1 and A2 allele, explore the impact of population size on drift?
• For a larger population size (1000) explore settings that reflect balancing selection.. (w11:.9; w12:1, w22=.95)
• What happens, if you decrease the population size?
• Using a small initial frequency of allele 1 (e.g. 0.01 with a population of 50) and a large fitness advantage for this allele, perform simulations for 20 populations.
  What does this suggest for the effectiveness of natural selection? Does natural selection acting on a single advantageous allele work better in a large population?
• What, if any, is the difference between a mutation and a substitution? (More than on answer is possible.)

Assignments for Friday

• Take-home exam #6 is due

For today:

Bootstrap - non-parametric (sampling with replacement)

Bootstrap - parametric (simulation using parameters extracted from a model)

Model parameters to define or estimate from the data:

• frequencies of nucleotides and aa.
• Shape parameter of Gamma distribution to describe among site rate variation. (one alternative: covarion model)
• type of substitution matrix

Else:

• Where should support values go if one draws a tree?
• Test which model is most appropriate.,
• Is a general time reversible model appropriate (is the composition of sequences homogeneous)?
• What data to use? (nucleotides, recoded as R/Y?, aa, recoded into Dayhof groups, codons (61x61 substitution matrix)
• Advantages and disadvantages to exclude positions with gaps.
• Programs frequently used phyml (seaview), Raxml, iQtree, MrBayes, fasttree

PPT Slides for today

Goals class 18-19:

• Know the names and characteristics of the different types of homology
• Know about the different ways the tree of life can be rooted.
• Know the reasons why a gene tree might be different from the species tree
• Be able to read bipartition tables
• Know the basic structure with which program in phylip work.
• Know how bootstrap samples are created
• Know the principle behind parsimony analysis and Occam's razor (or Ockham's razor, aka lex parsimoniae)
• Know the similarity and differences between parsimony and maximum likelihood based phylogenetic reconstruction.
• Know about the relationship between maximum likelihood, posterior and prior probability.