Computational resources: a regulatory need, a tool for research.

Background: We have recently reported a new method based on protein motifs to predict allergenicity by protein sequence. By using the more complete allergen database Allergome (www.allergome.org), as the basis for the motif identification process, we obtained a new set of allergenic motifs. We studied whether an allergenic motif may be used instead of a full length allergen for predicting allergenicity.

Methods: Sequence accession numbers of allergens were received from A. Mari of the Allergome Project. The sequences were downloaded and integrated in the allergen database, which was used to produce new motifs applicable for allergenicity prediction. The motif containing allergenic tropomyosins of different non vertebrate species was chosen. A consensus sequence of the tropomyosin motif was synthesized and expressed in Escherichia coli (E. coli). The purified peptide was analyzed for specificity with enzyme linked-immunosorbent assay (ELISA).

Results: Based on the Allergome dataset, we calculated 69 motifs containing 912 sequences. Prediction performance of this new set of allergenic motifs is increased compared to the original 52 motifs. The IgE immunological reactivity of the recombinant tropomyosin motif was comparable to that obtained with a full length recombinant shrimp tropomyosin (rPen a 1) tested in the ImmunoCAP system.

Discussion: The use of the motif-based allergenicity prediction method is more accurate than the guidelines proposed by FAO/WHO, however continuous updates of the motif dataset are necessary to guarantee a high prediction performance of this method. The experiments performed with the tropomyosin motif showed that peptides based on the consensus sequence of motifs may represent the allergenic epitope, which can be used in future diagnostic or therapeutic approaches.