This page will be used to post the answers to questions from students (mostly having to do with course content).
Send your questions to me (smount@umd.edu) and I will post your question and the answer here for all to see.
(These are in chronological order, most recent at the top).
Question 09.04.01.13
A student wrote (regarding question 1 on homework 3):
I had a question about finding paralogs of our yeast gene in C. elegans
(question 1 from homework 3).
(These genes are all orthologs of your yeast gene; they are paralogs of each other.)
I started my analysis by searching
the C. elegans refseq database. ... I took the protein from my top hit for my first search (yeast-elegans) and
performed a search against the C. elegans data base.
I compared the bit scores of my second search (elegans-elegans) to that of my
first (yeast-elegans) and selected as "paralogs" those hits (from search 2) which
had better bit scores then the highest hit from search 1.
Is this the correct method for finding paralogs?
Answer:
The approach is exactly right, but be sure to compare scores that were obtained using the same type of blast and the same parameters.
The criterion is:
Genes that encode proteins that are more similar to each other than they are to the protein encoded by your yeast gene, but more similar to your yeast gene than to any other yeast gene.
The simplest way to do this is to use your C. elegans protein, once it has been identified, as query to perform blastp against RefSeq from C. elegans and (without changing parameters) RefSeq from S. cerevisiae. You could also do one search against RefSeq proteins from all species and sort your results by taxonomy. This figure might be helpful:
If your assigned gene is YFG1, then I'm looking for YFL1. If your yeast gene is YFG2, then I'm looking for YFL2 and YFL3.
Question 09.03.31.22
Several students asked for clarification of question 2b in Homework 2:
"State which proteins you think are orthologous in each of the two examples."
Answer:
Example 1 is the LSM proteins. Example 2 is the kinases.
This is two questions:
For example 1, which proteins are orthologous?
For example 2, which proteins are orthologous?
The guide trees are just to help you think about the possibilities.
Tree C is the case where all Yeast-Fly pairs are orthologous.
Some of you were a bit confused by the three labelings, so I drew out all three trees, including a line at the time of the last common ancestor.
Question 09.03.30.01
A student wrote:
I'm having issues with Question 10, Homework 1 because I don't think I'm getting the results I'm supposed to get.
Answer:
Something changed at NCBI after I checked this question.
When I wrote it, the hit that I was looking for was just barely significant, but a little investigation showed that it is real.
Now, with the default settings, you get nothing.
Please change the settings as shown.
For question 8:
For question 10:
Question 09.03.29.24
A student wrote:
For the following question 12 how do you go about distinguishing if the pair is an ortholog or paralog, I think I understand the definition of the two however am not sure how to analyze them?
Answer:
I explained this in class. If two homologs diverged in a speciation event, they are orthologs. If they diverged in a gene duplication event, then they are paralogs. If you can draw a species tree and a gene tree, then it is straighforward to apply this principle. Trace the line between the two genes. Where did they diverge? In the cases that I am asking about in this homework assignment, application of the definition is straightforward because you are given the trees. If you have only sequences, then things can be ambiguous. We will revisit the issue in class, but you can do these problems by applying this principle. You may want to consult Fitch (2001) on homology (PDF, PMID 10782117) for a full explanation.
One minor point may be ambiguous: the time of divergence of delta globin. Please assume that the divergence of delta globin and beta globin (a range of 40-80 Myr. is given in the figure) was after the divergence of mouse and human (roughly 75 million years).
Question 09.03.29.23
A student wrote:
For both question sets 8-9 and 10-11 do you do the blast search using the accession numbers under the organism Homo sapiens?
Answer:
Yes; use homo sapiens. See below (09.03.29.03)
modified: March 29, 2009