Skip to main content
. 2008 Apr 29;36(10):3455–3462. doi: 10.1093/nar/gkn168

Figure 3.

Figure 3.

Performance comparison between the BLAST-end method (plotted points) and the PGA method (vertical bars) using different simulated data sets. (A) The assembly accuracies on data sets containing different numbers of simulated contigs of random fragments derived from Synechococcus sp. WH8102. The genomes of Synechococcus sp. CC9605 (S9605), Synechococcus sp. CC9902 (S9902) and P. marinus MIT 9313 (P9313) were used as the reference genomes, respectively. (B) The assembly accuracies on data sets containing different numbers of simulated contigs of random fragments derived from Shewanella sp. ANA-3. The genomes of S. oneidiensis (Sone), S. amazonensis (Sama), and S. frigidimarina (Sfri) were used as references, respectively. (C) The overall success rate for gap closure attained if one uses the four best predictions from the relaxed PGA method (PGA-extended, see text for detail) with the same data sets used in Figure 3A. (D) The overall success rate for gap closure attained using the four best predictions from the relaxed PGA method (PGA-extended, see text for detail) with the same data sets used in Figure 3B.