Methods predict candidate epitopes in a protein by: providing a conformational sampling engine; obtaining a model of a protein suitable for use with the conformational sampling engine and comprising a native structure; applying a collective coordinate bias to the protein model, the collective coordinate bias causing the conformational sampling engine to in turn force the protein model to at least partially unfold from its native structure to an updated structure, wherein the collective coordinate bias is applied globally to at least a substantial portion of the protein model and is impartial as to where, within the substantial portion of the protein model, unfolding occurs; and analyzing the updated structure to identify one or more candidate epitopes, where the one or more candidate epitopes exhibit indicia of localized unfolding.