Skip to main content
Figure 2 | BMC Bioinformatics

Figure 2

From: Evolutionary rates at codon sites may be used to align sequences and infer protein domain function

Figure 2

FIRE, T-Coffee, ClustalW and MAFFT MSAs. The alignments generated by (A) FATCAT, (B) FIRE, (C) T-Coffee, (D) ClustalW and (E) MAFFT algorithms for kappa and lambda antibody variable regions (data set 6 in Table 1) are displayed. Only sequences corresponding to the two structure files in the FATCAT alignment and the representative sequences from the two clades aligned by FIRE are shown for each of the other three MSAs. Using the FATCAT alignment as an independent standard-of-truth reference, correctly aligned residue pairs in the other four MSAs were identified (shaded regions). Overall, T-Coffee and MAFFT produced the most accurate alignments, however, FIRE performed better than ClustalW demonstrating the viability of using an evolutionary rates-based approach to sequence analysis when sequence similarity is low. In addition, the short stretches of conserved amino acids (indicated by +) inflate the performances of the three homology-based methods relative to FIRE (see text for discussion).

Back to article page