Dear Aparna,

It really depends on what you are interested with. For example, if you are studying a membrane protein such as an ion channel with transmembrane domains, they it would be wise to choose an extracellular segment or a cysteine rich intracellular domain rather than a TM domain. Also you should check whether the segment (~18 aa) that you picked is not identical to another protein. In other words you need pick a segment that is not shared by other proteins or similar proteins.

When I did my PhD I raised my own antibodies in rabbits and characterzied them using various techniques. To increase antigenicity I combined my segment of amino acids with KLH. For details and inspiration you may want to download my PhD and resultant publication in J Comp Neurol that have more than 200 citations.

best wishes, Refik

Thesis Molecular physiology of ATP-gated ion channels assembled fro...

Article Distribution of the P2X2 receptor subunit of the ATP-gated i...

Dear Aparna,

Usually exposed (solvent accessible) domains are good for generating antibodies.

For IHC you usually need a mAb against a linear epitope.

There are several way to predict which peptide sequence is a good B cell epitope:

http://tools.iedb.org/main/bcell/

http://www.imtech.res.in/raghava/lbtope/

http://ailab.ist.psu.edu/bcpred/

http://www.cbs.dtu.dk/services/BepiPred/

http://sysbio.unl.edu/SVMTriP/

http://omictools.com/b-cell-epitopes-category

Best,

Piero