Skip to main content
Figure 4 | BMC Bioinformatics

Figure 4

From: Improving pairwise sequence alignment accuracy using near-optimal protein sequence alignments

Figure 4

Boxplot depicting the distribution of shift scores for structural alignments and near-optimal sequence alignments. Shift scores are calculated relative to the Dali alignment for each pair of proteins. Shift scores, which range from -0.2 to 1.0, were plotted on a logarithmic scale to highlight the region between 0.6 and 1.0. DALI shift scores (the reference) are plotted as blue X's, CE as green squares, Matras as magenta +'s, and LSQMAN as red triangles. The Zuker near-optimal alignments are those with a similarity score within 75% of the optimal score. The boxplot for 1000 probA alignments is plotted to the right of the Zuker boxplots for each pair. The three panels present the groups of proteins that share high, medium, and low statistically significant sequence similarity. Protein pairs within a similarity group are ordered from left to right by expectation value (i.e. most similar to least similar). Pairs 25 and 42 from Figures 1 & 2 are highlighted with a darker blue rectangle. The colored dots at the bottom of each column represent either very poor or nonexistent alignments produced by the structural alignment programs. The colors match those of the symbols for the high quality alignments.

Back to article page