Immunology Prac 3 Bioinformatics
Immunology Prac 3. Bioinformatics
For your write-up of this practical, give an introduction on computational methods of epitope prediction (approx half a page), noting both T and B cell, and the differences. Answer the questions posed below in bold, with reasoning. You can answer within this document.
Introduction.
Write your introduction here.
Practical analyses.
1. Use the following sequence:
>gi|45384056|ref|NP_990483.1| ovalbumin [Gallus gallus]
MGSIGAASMEFCFDVFKELKVHHANENIFYCPIAIMSALAMVYLGAKDSTRTQINKVVRFDKLPGFGDSIEAQCGTSVNVHSSLRDILNQITKPNDVYSFSLASRLYAEERYPILPEYLQCVKELYRGGLEPINFQTAADQARELINSWVESQTNGIIRNVLQPSSVDSQTAMVLVNAIVFKGLWEKTFKDEDTQAMPFRVTEQESKPVQMMYQIGLFRVASMASEKMKILELPFASGTMSMLVLLPDEVSGLEQLESIINFEKLTEWTSSNVMEERKIKVYLPRMKMEEKYNLTSVLMAMGITDVFSSSANLSGISSAESLKISQAVHAAHAEINEAGREVVGSAEAGVDAASVSEEFRADHPFLFCIKHIATNAVLFFGRCVSP
Using http://www.syfpeithi.de/, determine which is the dominant CD8 epitope in mice. Click Epitope prediction. Use H2-kb, and search for octomers.
Try human HLA-B*08, again octomers. Is the same peptide dominant? Is it predicted that HLA-B*08 can present this peptide?
This peptide is widely used in antigen presentation assays in immunology.
Does another epitope predictor http://tools.immuneepitope.org/mhci/ also predict this as the dominant peptide in H2-kb mice?
Using the Hopp-Woods tool at https://web.expasy.org/protscale/ find the most hydrophilic region of the protein. Use the default window size (9). Where is it? Why might it be important to know this?
2. If you were going to design a peptide vaccine to (A) induce CTL’s, or (B) induce antibodies from the following sequence, which regions would you choose? Assume you will test in H2-Db mice.
LPKSFDARVEWPHCPSISEIRDQSSCGSCWAFGAVEAMSDRICIKSKGKHKPFLSAENLVSCCSSCGMGCNGGFPHSAWLYWKNQGIVTGDLYNTTNGCQPYEFPPCEHHVIGPLPSCDGDVETPSCKTNCQPGYNIPYEKD
3. B cell epitope prediction.
These are much more complex, as often epitopes represented by antibody are not linear, and therefore are derived from different regions of the polypeptide chain. If detailed knowledge of the protein structure is known, then prediction may be easier.
Generally, prediction revolves around the prediction of surface regions.
Go to http://tools.immuneepitope.org/main/bcell/
Have a look at the various tools available. Note that all except one rely of some structural information. In this prac we will look at the first two, prediction of linearar epitopes using various tools and using Discotope to use structural information for prediction.
A. Prediction of linear epitopes from protein sequence.
Input the ovalbumin sequence.
Use each of the 7 algorithms to see the output. Paste the summary of each in your report.
What does each predict for the ovalbumin sequence? Are they similar?
Explain why each of the 7 prediction tools is important?
Are any of the peptides predicted by more than one method?
B. Discotope - Prediction of epitopes from protein structure
Enter 1OVA into the PDB ID box. This is the entry in the Protein Data Bank for ovalbumin structure. Enter A for the chain. Leave the version at 1.1. Run the tool. Paste the output here. What is your interpretation? Does it concur with what you found using the Hopp-Woods tool? Why or why not?
