(交流）PREDICTION OF B CELL EPITOPE （外国专家给我的信）

小弟因为课题需要，要对融合蛋白进行B CELL EPITOPE 预测，一时没有办法下手。在网站上游荡许久，找到一个专门从事这项研究的专家，于是给他发信求助。他还真是好人，给我回了这封信，关于表位预测的，希望对大家有用：

Thank you for your interest in my work. Unfortunately there is no accurate
method for predicting Linear B cell epitopes currently available via the
internet. Protein sequence profiling methods using Hydrophilicity, Flexibility,
and Accessibility etc... do not work well. However if you choose to use one of
these methods they can be found here: http://www.imtech.res.in/raghava/bcepred/

If your intention is to successfully predict only 1 or 2 epitopes form your
protein I would recommend trying to find protruding loop or turn structures
located on the surface of the protein. If there is a known 3D structure of your
protein in the PDB database this should be quite easy. Programs such as DSSP or
STRIDE can analyze your 3D protein structure and tell you were these regions
are and how accessible they are.
Available here:
http://bioweb.pasteur.fr/seqanal/interfaces/dssp-simple.html
http://bioweb.pasteur.fr/seqanal/interfaces/stride.html

If you do not have a known structure for your protein I would use a secondary
structure prediction program such as JPRED or PREDATOR to predict the location
of loop and turn structures within your protein

http://www.compbio.dundee.ac.uk/~www-jpred/
http://bioweb.pasteur.fr/seqanal/interfaces/predator-simple.html

Once you have found a region of your protein that you think is likely to
contain a B cell epitope I would also recommend synthesising a large peptide of
approximately 15 amino acids to cover it as the local secondary structure maybe
important in enabling an antibody to cross-react.

Finally, I do not know a great deal about recombinant protein technology but I
would imagine that so long as a predicted B cell epitope is still accessible
and is not masked from being fussed to another protein (i.e. The predicted
epitope is not facing the join to the other protein) the fussing or your
protein should not have a significant effect.

I hope I have been of some help to you,