Need to look at structure of the virus.  Things on the outside of the virus (coat or other attachment on surface) normally good candidates.  If you know the gene sequence - several sites on the net that you can use to predicts antigenic sequences.  Have used these for a few viruses - all the peptides generated usining these tools produced very good immune responses. Peptides were produced chemically as MAP's for immunization.

It's not my area of expertise, but I would think that a viral protein found on the outside of the virus would be a better antigen for a vaccine than one found on the inside, like polymerase.

Thanks for yours answers. T- and B-cell epitope combination is a trend. http://www.ncbi.nlm.nih.gov/pubmed/22044118

T cell epitopes can be found on either the outside or inside of a virus. The classical example is the nucleoprotein 118 epitope in LCVM (inside) in the Balb/c mouse as well as the nucleoprotein 396 (inside) & glycoprotein 33 (potentially outside) in the B6 mouse. T cell epitopes, because the are processed before presentation, can be found anywhere in the proteome. On the other hand B cell epitopes are almost always on the outside of the virus and for the most part are hydrophilic in nature so as to be accessible for the antibodies. An other difference is that while T cell epitopes are linear in nature and sequential in the protein structure, B cell epitopes CAN be linear in the protein OR a combination of different residues in different 'loops' of the protein and even a combination of residues from different proteins, i.e. these epitopes CAN be 3D in nature and lost in conformation changes and denaturation of the virus.

As for the original question. NO quantity is NOT always a good indicator for a vaccine candidate, it obviously critically depends on the presence of an epitope and the quality of that epitope, i.e. it's binding affinity to the intended MHC and the presence of an T cell precursor population with appropriate binding affinity or precursor Ab population. In addition, for most anti-body response based vaccines the epitope selected would need to be neutralizing (i.e. binding of the Ab would need to inhibit binding and/or entry of the virus into the cell).

A book could be, and probably has been, written on this subject and even then we don't know all the parameters that go into making a 'good' vaccine. Evidence the failure to produce an effective vaccine for AIDS and Flu as well as a number of other important diseases which don't have the particular 'mutation problem' these have.

yes i agree with @Edmund Pool  to prepare vaccine the protein must contain antigenic sequences which would be checked by only expression method.

Thanks for your answers. OK, I understood that there are many additional and possibly more important factors to take into account than viral protein expression.