DoMosaic - Analysis of the Mosaic-like domain arrangements in proteins

Sequence analysis is widely used to infer function from one protein sequence to another. One of the remaining hurdles in pushing further the limits of efficient usage is the modular and mobile architecture of proteins. Although many resources provide precompiled signatures, they are not yet used in conjunction with pairwise comparisons to investigate the modular architecture and phylogeny. We present a program, doMosaic, which combines existing domain definitions with an adapted sequence alignment algorithm. It is based on the Smith-Waterman scheme, can be combined with trees derived from the domains and provides a user-friendly graphical interface. The method enables fast and efficient analysis of domain duplication events within one or in-between two sequences. Therefore, it enables refined functional annotation.