To all,
I am in the process of determining several antigenic sites from protein of
interest on in order to raise rabbit polyclonal antibodies by immunising the
animals with selected 15-mer amino acids peptides. I have used algorythms such
as Hopp and Wood, Kyte and Doolittle, and so on, to make my choice. But some
of my readings have been quite pessimistics about this approach arguing that
discontinued epitopes are more likely to be antigenic than continuous epitopes
and that use of informatic algorythms is about as good as a structured wild
guess...
My question: is it possible to get a combination of amino acids that would be
more likely to mimics the native protein?
I have in mind the presence of couple of cystein in the peptide that would
make disulfide bound getting the peptide to adopt a non-flat conformation or
the presence of proline within the sequence.
Somebody have any ideas?
TIA,
Yves Blais
--
