Are you looking for antibodies that crossreact with the folded protein? In this case you look for antigen peptides that are accessible on the protein surface and somewhat conformationally flexible, that is, loop and turn regions on the protein. There are a number of sequence based methods that highlight potential epitopes, http://tools.immuneepitope.org/bcell/ , http://ailab.ist.psu.edu/bcpred/predict.html, http://curie.utmb.edu/B-Cell.html, http://www.cbs.dtu.dk/services/BepiPred/, http://sysbio.unl.edu/SVMTriP/prediction.php . 

Are you looking for antibodies that crossreact with the folded protein? In this case you look for antigen peptides that are accessible on the protein surface and somewhat conformationally flexible, that is, loop and turn regions on the protein. There are a number of sequence based methods that highlight potential epitopes, http://tools.immuneepitope.org/bcell/ , http://ailab.ist.psu.edu/bcpred/predict.html, http://curie.utmb.edu/B-Cell.html, http://www.cbs.dtu.dk/services/BepiPred/, http://sysbio.unl.edu/SVMTriP/prediction.php . 

Which immune response are you looking for?

1. If you are looking for a cellular rather than humoral response there are many bioinformatic tools that allow to predict T-cell epitopes [MHC I binding, MHC I processing, MHC I immunogenicity; see http://www.iedb.org/]

2. On the other hand, if you are looking for a humoral response, you could predict linear o discontinuos B-cell epitopes based on AA sequences [http://tools.iedb.org/bcell/] or 3D structures of proteins in PDB format [see http://tools.iedb.org/discotope/ and http://tools.iedb.org/ellipro/], respectively.

3. If antisera obtained from an infected host are available, linear B-cell epitopes can also be identified by peptide array [http://www.pepperprint.com/leading-peptide-microarray-solutions/]

I agree with the comments given above. You need to conduct MDS study to predict the peptide epitope or antigens.