Table 1: Contingency table for the distribution of replication-related effects on the GC-skew and AT-skew in Bacteria and Archaea. The columns of the table indicate the sign of the nucleotide skews on the leading strand, induced by the replication mechanism. For instance, G > C means that the leading strand is enriched in G over C, thus the sign of the GC-skew on the leading strand is positive, while G = C means that no significant effect of replication on the GC-skew could be detected. The cells of the table contain the number of chromosomes that we found to present the respective GC-skew and AT-skew signs on the leading strand.
Table 2: Comparison between our approach and the methods proposed by Tillier and Collins, 2000 and Lobry and Sueoka, 2002, for 173 chromosomes for which we have performed significance computations. Cases where the p-values are not available for these two methods correspond to chromosomes where the origin and terminus could not be determined with certainty.
Figure 1:
![]()
Figure 1: Results from simulations where the effect of replication is constant (r=0.1), the standard variation of the nucleotide skews is constant for both groups of genes (sd=0.01), the breakpoint location is constant (fr=0.25), and the coding sequence related effects (m) vary. The y-axis represents the distance between the real and the estimated breakpoints.
Figure 2:
![]()
Figure 2: Results from simulations where the coding-sequence related effect is constant (m=0.1), the standard variation of the nucleotide skews is constant for both groups of genes (sd=0.01), the breakpoint location is constant (fr=0.25), and the replication effect (r) varies. The y-axis represents the distance between the real and the estimated breakpoints.
Figure 3:
![]()
Figure 3: Results from simulations where the replication and the coding-sequence related effects are constant (m=2.0 and r=0.1), the standard variation of the nucleotide skews is constant for both groups of genes (sd=0.01), but the position of the breakpoint (fr) varies. The y-axis represents the distance between the real and the estimated breakpoints.
Figure 4:
![]()
![]()
Figure 4: Results from simulations where the effect of replication is constant (r=0.01 for the first graph and r=0.1 for the second graph), the coding sequence related effect is constant (m=2.0), the breakpoint location is constant (fr=0.25) and the standard variation of the nucleotide skews varies. The y-axis represents the distance between the real and the estimated breakpoints. This document was translated from LATEX by HEVEA.