Scanning the sequence databank with an alignment of sequences is a sensitive method to detect weakly similar sequences belonging to the same family [39][25]. A scan using the alignment in Figure 1 ranked the E. coli protein diadenosine tetra-phosphatase [40] within the low scoring ``tail'' of known phosphatase sequences. A scan with the diadenosine tetra-phosphatase sequence showed the lambda protein phosphatase and Drosophila melanogaster PPY to be the most similar proteins in the databank, followed by the Trypanosoma brucei PP1.
Figure 2 illustrates an alignment of the E. coli diadenosine
tetra-phosphatase with the protein phosphatases from 80, and
lambda, and a representative set of phosphatases from Figure 1. The
80, lambda and tetra-phosphatase sequences share greatest
similarity with eukaryotic protein phosphatases in the first 109
positions of the alignment which are shown in Figure 2. The
similarity following position 109 is very low and sequence alignments
in this region are ambiguous.
The most striking regions of conservation between the sequences in
Figure 2 start at positions 34, 64 and 96 (58, 87 and 137 on Figure 1). These similarities suggest that the E. coli diadenosine
tetra-phosphatase, 80 and eukaryotic protein phosphatases have
diverged from a common evolutionary ancestor.