Accuracy of alignments generated using different methods on the HOMEP2 data set.

Results are sorted according to the level of sequence similarity of the sequence pair, in percentage identity. The number of pairwise alignments is shown in parentheses. The percentage of correctly aligned residues (% correct) and average shift error size (shift) with respect to the structure-based reference alignments (see Methods) are reported. *Values marked with an asterisk in this and all other tables are not significantly different from those of AlignMePST (p-value >0.05) based on a pairwise Wilcoxon signed rank test. All other values are significantly different from those of AlignMePST. Entries in bold in this table, and all subsequent tables, indicate the highest or best scores in that column, including all values that are not significantly different from the best scores.