Darien,

several options: 1) paste the sequence of your protein into e.g. the SYFPEITHI algorithm and search for likely epitope candidates. Then have them synthesized and tested.

2) If the literature already proposes immundominant epitopes, have the synthesized and tested.

3) if you would like to be comprehensive, purchase a library of peptides, e.g. 15mers overlapping by 4 amino acids, that span the whole protein. Test this library by creating different peptide pools that allow you to dissect the immune response.

(The most comprehensive but also most expensive strategy.) For CD4 epitopes you may need to use longer peptides.

Read-out could be intracellular cytokine staining or ELISPOT .

Matthias

Another resource with some robust epitope prediction tools is the IEDB website.

For HLA class II 15mers are usually long enough, at least for foreign proteins. I believe Joan Goverman did publish about an MS epitope with a minimum length of 18, implying that minimal epitopes can be unusually long for some self-epitopes.