Events News Research CBS CBS Publications Bioinformatics
Staff Contact About Internal CBS CBS Other

Exercise: Identifying immunologically relevant target proteins using PathogenFinder

Mette Voldby Larsen (metteb@cbs.dtu.dk)


Overview

Not all proteins are equally relevant as immunological targets. In general, proteins that are located to the inside of bacteria cannot be seen by the immune system. Furthermore, some proteins are present in innocuous as well as pathogenic bacteria.

Many different approaches can be imaginged for selecting immunologically relevant proteins. In this exercise, we will use the PathogenFinder service. This service aims at predicting whether an uploaded genome or proteome is from a pathogenic bacterium or not. As a bonus, the subset of proteins that formed the basis of the prediction is listed.

Please note: The PathogenFinder service is under development and not fully documented.


Use the PathogenFinder service to predict whether or not the proteome you ended up with in the previous exercise (Predicting genes using Prodigal) is likely to be from a pathogen or not.

When the prediction is finished, you will see some information concerning the probability of the uploaded proteome being from a human pathogen along with some other statistics. Below the statistics is a list of the proteins that PathogenFinder based its prediction on. "Input Sequence" is equivalent to the headers of the proteins predicted by Prodigal. The lines marked "Matched Family", contain information concerning the matched protein families from the database used by PathogenFinder. The first four columns concerns the isolate from which the protein that represents the family originates. The last three columns concerns the protein representing the family (description, accession number and how identical the protein in the proteom you inputted is to the protein in PathogenFinder's database).

Which of the proteins could be relevant as immunological targets? If you don't have any suggesstions, you might be inspired by the article concerning Shiga toxin antibody therapy.

As a next step, you could predict B-cell epitopes (yesterday's topic) and MHC class II epitopes (which you will hear about next week) in the proteins, but that is beyond the scope of this exercise. So instead ... Have a nice lunch break!