Computational selection of novel antigenic targets in the Mycobacterium bovis proteome

The discovery of novel antigens is an essential requirement in devising new diagnostics for use in both M. tuberculosis (Mtb) and M. bovis control programmes. Reverse vaccinology is now a feasible method of extracting potential immunogenic epitopes from bacterial genomes to reduce the cost of experimental screening of antigens for anamnestic responses in infected hosts. Since a significant focus has been on the role of CD4+ T cells, the ability to predict peptide binding to MHC-II molecules is seen as a key step in discovery.Previous antigen-mining experiments for identification of novel diagnostic or vaccine candidates for human and bovine TB follow a targeted approach, where specific groups of proteins suspected to contain likely candidates are identified and evaluated for  mmunogenicity. A disadvantage of those approaches is that they are restricted to a relatively small set of proteins biased by the initial selection criteria. Our objective was to computationally select antigens in a less biased manner.