Peptide design

Our immunizing peptides consist of 10-15 residues. This is large enough to elicit a sufficient immune response without losing specificity. It is also short enough to keep the number of epitopes limited and again to keep the resulting antibody as specific to the target protein as a monoclonal antibody would be.

The choice of the immunizing peptide is not random. Many parts of a protein will not be fit at all to generate a successful antibody. Certain combinations of amino acids are simply not favourable. One needs to pick a peptide from a part of protein that is likely situated at the surface of the native folded protein and that this part did not undergo any post-translational modifications, or interactions. This way, the resulting antibody may be better fit for applications such as IHC, IF and quantitative analysis in an ELISA-type platform.

We use an automated system to apply seven commonly known algorithms (two for hydrophobicity, two for antigenicity, two for protein surface seeking sequences and one for secondary structures) in order to generate a selection of candidate peptides that would meet the criteria as described above. This automated system is part of our intellectual proprietary. Then the candidates with the highest scores will be scrutinized further by checking both NCBI and UniprotKB in order to see which of the candidate peptides best serves its purpose. 

The final peptide sequence goes through a blastp search in the mammalian protein database of NCBI. When the peptide sequence shows overwhelming overlap with other species, we will report such species as expected to cross-react with the resulting antibody. We limit the species we report on our datasheets to human, mouse, rat, dog, pig and cow – though often similarity to many other species are found.

                                                                                                                                                         next>