抄録
A new computational system was developed, which find high quality motif that can discriminates the unique protein family without false entries. In this system, motif patterns were plotted to the dispersion diagram between observed and expected motif numbers in SWISSPROT database. It was revealed that the motifs with different quality could be clearly assigned on different areas by discrimination line due to the experimental assignment in PROSITE database, together with the alignment analysis of each members of family, relating motifs. The result suggested that the evaluation of motif is feasible using the relation between the observed and the expected of motif match and the similarity scores of family proteins.
