|
|
This optional lab contains a problem dealing with modelling. The exercise uses some of the web-based tools discussed in lecture. The lab assumes that you will return to earlier labs if you need help with a given web site that you have used before. A listing of potentially useful web sites is given at the end of the text.
Lab 1 help guide Lab 2 help guide Lab 3 help guide
Modelling In this example you will obtain structural information from a cloned sequence. A molecular biologist in your group has found a nucleic acid sequence that they think is interesting. They want to know the function of that sequence and if possible, they would like to know any structural information so they can interpret some mutagenesis results on the expressed gene.
The marker sequence they have provided you is this:
tccaacagctctgacctgggcagccttcctgagaaaa
1. Is this marker sequence likely human? (Hint: Perform a BLAT search) What gene is it in? Get the protein sequence of the expressed gene. Questions to think about: Why isn't there a known gene listed for this sequence in the genome browser?
2. What are the likely function and likely homologs for this protein?
3. Find a close homolog (or two) with an X-ray crystal structure and a more distant homolog.
4. Using whatever method you like, align the homologs to your sequence. Are your homologs and this gene likely related?
5. Does a secondary structure prediction confirm your hypothesis that your homologs and this protein are related? How about for the more distantly related sequence?
6. Using one of the homologs, build a model of your sequence.
Building a model of a protein at the Swiss Model site requires the an email address in order to receive your results. Since one has NOT been supplied, this part of the exercise will be simulated.
The following sequence was pasted into the Swiss Model form resulting in the return of this email message. The email header were removed from the message to leave the following PDB file. Save the final file to your local machine.
7. Evaluate the model using common model evaluation programs. (http://biotech.ebi.ac.uk:8400/)
Use the Quick start link to start the process. On the next page, use File Upload to browse your local machine and upload the file to the server. Depending on your patience level, you can check your model with either a quick, medium or long run via the Select checks to run link. Reload the results of your run until the process says that all the checks have been completed. Don't be surprised if some of the checks fail. Be sure to remove your data from the server when you are finished looking at your results.
8. What do you expect is the quality of your model?
If a researcher approaches you about possible zinc binding activity in your protein, do you think you can predict where the zinc might be binding? Conversely, if a researcher approaches you about the conformation of two loops that are important to function, do you think your models might be useful? Why or why not?