Abstract
Multiple sequence alignment (MSA) is a method to find the relationship between biological sequences. A Domain is an evolutionary unit in protein. Generally, the proteins contains similar domains are seen as homologous. Because domains have such properties, they are worth of being taken into account when doing alignments. However, most MSA tools ignore the information about domains. In this paper, a domain-based multiple sequence alignments tool – DOBALI is introduced. DOBALI performs preprocesses before doing MSA. It lets users assign the positions of domains. Then, the possible domains are detected, and the segmented domains are combined and rearranged. After that, all the same kinds of domains are aligned together by an existing MSA tool chosen by the users. The result shows the alignments are better than which produced by none-preprocessing in some conditions.