homework 3 done

.docx

School

California State University, Los Angeles *

*We aren’t endorsed by this school

Course

4180

Subject

Biology

Date

Oct 30, 2023

Type

docx

Pages

Uploaded by MegaBravery6485

BIOL 4180 Homework assignment #3 Due Oct 16 Name Lorenzo Angelus Fernandez See new document posted to Week 7 module, “ 1_instructions, 2023 ” for detailed directions on how to perform a ML analysis using RAxML on a personal computer versus on the CIPRES web platform, and how to visualize a tree using MEGA. 1] Download the zip file “Cyerce COI RAxML for 4180” posted under the ‘Links & Resources” header for Canvas Week 7. This is the output for the partitioned RAxML analysis we did in class using CIPRES. You can unzip the file to see the 16 files generated by the analysis. Using a text editor, you can also open and view the output files to find some of the answers below; lecture slides should help point you to the right files. (a) What were the parameters of the two models fit to the data during the analysis? Was anything surprising about the parameter values to you, considering what we discussed in class? (2.5 pt) The model was very complex, RAxML worked hard to give the tree with high but not significant bootstrap values. GTR’s model had less parameters and gave higher bootstrap values that were significant across the whole tree. More complex models tend to have higher amounts of significance, but in this case the data wasn’t presented accurately by the more complex model. (b) How long did the search for the maximum likelihood (best-fit) tree itself take? How long did the bootstrapping portion of the analysis take? (1 pt) My computer took 10 to 15 minutes to make each of the trees, but getting it actually running took far longer because RAxML was giving a lot of trouble and would crash often. It would overheat my computer and even cause it to power down. This means I had to perform the analysis many times and take pictures because screenshotting took too long. This caused it to take MUCH longer than just the raw analysis. (c) In MEGA, FigTree or an equivalent program, plot the ML tree showing only significant support values. To make the tree simpler to see, try to collapse each species. MEGA will let you type in the name for any group you collapse; you can use the full name (for instance, “cristallina”) or just the first few letters (“cris”, “ant”) for each. When I say species, I mean each group where the samples have the same name: cristallina, antillensis, nicholasi, etc. There are seven on this tree. (2 pt) /25 1 1

(d) Compare this tree to the neighbor-joining tree for the same dataset from Homework 2. If you need to, just create that NJ tree again using 100 bootstraps; the input file was “Cyerce COI MEGA” under the Assignment for Week 5. What species form clades on the ML tree that were not supported as clades in the NJ analysis? (2.5 pt) 2 2

The browneveorum and willeteorum species have less significance in homework 2’s tree. The bootstrap values however are higher in the older analysis yet with a node value of 97 in this test but 74 in the past analysis. 2] Below is the code used in the partitioning file to run two models for COI, one positions 1+2 combined, and a second model for the 3 rd codon position: DNA, COIcodon1andcodon2 = 1-658\3, 3-658\3 DNA, COIcodon3 = 2-658\3 Now, imagine you are given a datafile that starts with 442 base-pairs (bp) of the 16S gene, followed by 658 bp of COI as a concatenated (combined) alignment. Rewrite this block of code to (a) use one model for all of the 16S partition; (b) a second model for codon positions 1+2 of COI combined; and (c) a third model for the 3 rd codon position of COI. (3 pt) DNA, 16S = 1-442 DNA, COIcodon1andcodon2 = 443-658\3 DNA, COIcodon3 = 444-658\3 3] For this question, please perform your own ML analysis using RAxML, with the input file “Cyerce_COI_RAxML”. This contains the same COI sequences you analyzed before in MEGA. You can do this analysis either using RAxML on your computer, or else through the CIPRES web portal—that is, the two ways we did ML analyses in class. To make this analysis different from the downloaded file: (a) run it without any partitions by codon position; that is, treat all codon positions as evolving under the same model of sequence evolution. (b) use one GTR+gamma model (c) use 100 bootstrap replications for your analysis (to make it go fast) (d) hide support values under 70% (e) root the tree on C. cristallina Insert or attach your ML tree, either by outputting as a graphic or taking a screenshot; you can collapse some of the species to make it easier to fit. Tell which program you used. (8 pt) 8 pts 3 3

Your preview ends here

Eager to read complete document? Join bartleby learn and gain access to the full version