Figure 1
From: Species-specific analysis of protein sequence motifs using mutual information
Schematic multiple alignment with m fragments of length n belonging to k user defined groups. During the MI-profile-calculation used vectors X i and Y are marked in red, whereas X i behaves like a sliding window from left to right (i = 1...n). X = (X1,..., X i ,..., X n ) and Y are derived by applying the enlarged Prosite motif PDOC00028 as a search-pattern to user-defined groups of protein sequences. By selecting fragments corresponding to the motiv a quasi multiple alignment was formed. In multitask the vector Y is formed from the user-given group labels. Here, to get a general idea, the alignment is colored and arranged per group-label of the vector Y. For demonstration conserved positions within the alignment (corresponding to the motif) are darkend. In case of pairwise alignments, only 2 groups would be stated.