Assignment4 Instructions
User Manual:
Open the PDF directly: View PDF 
.
Page Count: 1
Molecular Phylogenetics (EEOB 563)
Homework #4: Distance Matrix Methods
Part I. Distance methods using paper and pencil
J
.
Table 1:
Dog ATG ACC AAC ATT CGA AAA ACC CAC CCA CTA
Cat ATG ACC AAC ATT CGA AAA TCA CAC CCC CTT
Mouse ATG ACA AAC ATA CGA AAA ACA CAC CCA TTA
Pig ATG ACC AAC ATC CGA AAA TCA CAC CCA CTA
Human ATG ACC CCA ATA CGC AAA ACT AAC CCC CTA
1. Table 1 shows the first 30 bp of the mitochondrial cytochrome b gene for five mammals. Find
the Jukes-Cantor distance for each pair of species (I talked briefly about different models of
DNA evolution on Thursday, but here is the formula again: d = -3/4*ln(1-(4/3)p), were p is
uncorrected (observed) dissimilarity.)
2. Apply the algorithm we used in class to find the UPGMA tree based on JC distances. Draw
this tree and indicate all branch lengths.
3. Apply the neighbor-joining (NJ) algorithm to find the NJ tree. Draw this tree and indicate
all branch lengths.
Part II. Distance methods in PHYLIP.
You may want to check the PHYLIP documentation on
http://evolution.genetics.washington.edu/phylip and/or lab3 tutorial before doing this part.
Create a multiple sequence alignment using complete mitochondrial cytochrome b sequences
from the cob_nt.fasta file we used in class and save it in the PHYLIP format.
4. Calculate 4 different matrices using 4 models available in dnadist and perform a NJ analysis
with each of them. Do not submit these trees. Instead, calculate a strict consensus tree. Root
your consensus trees using an appropriate outgroup and include it with the rest of the
assignment.
5. Create 200 bootsrtap replicates of your data file and build a NJ tree for each of them (you
should choose one of the four models of nucleotide substitutions you used in #4).
Build a majority rule consensus tree for these trees, print it out, and submit with the rest of
the assignment.
Part III (Bonus point). FastME.
8. Use one of the distance matrices you calculated in #4 to calculate NJ trees with/without the
tree refinement with NNI and SPR. Did NNI/SPR post-processing changed the resulting tree
phylogeny? Which of the trees corresponds better to your understanding of mammalian
evolution? J
