Lab Entry 7 – Geneious Part II
Using the sequences acquired from the last lab entry (30 sequences in total), phylogenetic trees were created using various programs within Geneious. The first step to creating the trees was to align all of the sequences and then ‘clean’ them so that all of the sequences begin and end at the same point.
The next task was to create a model of molecular evolution using a program called jModelTest. I was unable to utilize this program because I use Windows/PC rather than MAC. Because of this, I observed Ryan Aquino go through the steps which began with placing the aligned sequence in the jModelTest sequence. The program runs going through 88 models and calculates likelihoods for each of those models. Once the program is done running through the various models, the best model is chosen using two models. The first is the Akaike Information Criterion and Bayesian Information Criterion which I was also unable to perform without the jModelTest.
I could run the MrBayes program within Geneious which was the next task to create a phylogenetic tree. Using the same aligned sequence that would’ve been used for the jModelTest, the sequence was run through the MrBayes program with specific parameters provided. The result was a tree, as well as two other graphs. These graphs provided indication that the program was not run long enough and I was instructed to run the program with different parameters that would allow the program to run for a longer duration. This resulted in a very different tree that was much more accurate based on comparing the species used in the alignment. The longer MrBayes run was performed after the lab session when more time was available.
The last step of the lab was to run programs for the maximum likelihood. There were two ways to do this: RAxML and PHYML. My computer would not run PHYML but I could run the RAxML program. The result was a tree different than the one provided by MrBayes. This tree was then run through a program to make a consensus tree. The RAxML tree was less accurate in comparison to the MrBayes tree. The timeliness of each of the programs displays the quality of the tree being produced. RAxML was very quick and showed some accuracy with relationships of species but the MrBayes program took much longer and was clearly more accurate when comparing species to one another and how they relate to each other.
