Fetch the two sequences using the expasy server or SRS.
Compare the two sequences using dotter or dottlet.
Extract every domain segment and make all the possible pairwise alignments.
- When you extract, be generous with the boundaries. It is easy to remove extra residues later.
- Do not forget to visually inspect your sequences for low complexity regions.This may guide you in the choice of the regions to extract.
In order to improve the alignment add one extra homologous segment. Compare the alignment with two sequences to the one made with three sequences.
Use Blast to fetch extra sequences homologous with yours. Be careful to extract only the relevant portions
Make a multiple sequence alignment and identify key residues.
Compare all the available methods on this sequence set. Do you see any difference?
Use SwissProt and Prosite annotation to propose a function for this domain?