Fig 1.
The proportion of DNMs in each mutational category in the three datasets.
CpG X>Y is an X>Y DNM at a CpG site, non X>Y is an X>Y DNM at a non-CpG site.
Fig 2.
Gamma distributions fitted to the DNM density at the 1MB scale.
In order of decreasing variance: Maroon–Wong, Blue–Francioli, Olive–Jonsson.
Fig 3.
Gamma distribution fitted at different scales and to different categories of mutation.
The gamma distribution fitted to the number of DNMs per window at different scales: 10KB (blue) 100KB (maroon), 1MB (olive) and 10MB (green) for the A) Francioli C) Wong, and E) Jonsson data; and the CV of the distribution fitted to various mutational categories at the 1MB scale for B) Francioli D) Wong, and F) Jonsson data.
Fig 4.
The coefficient of variation of the fitted distribution across scales.
Note both variables are plotted on a log-scale.
Table 1.
The correlation between different mutational types at the 1MB scale.
Fig 5.
The equilibrium GC content from a simulation of sequence evolution.
The equilibrium GC content from a simulation of sequence evolution is plotted against the current GC-content of the windows from which the mutation pattern was estimated. Note several of the points are coincident.
Table 2.
Correlation between the density of DNMs and the mutation rate estimates from the models of Aggarwala et al. [40] and Michaelson et al. [5] at the 1MB scale.
Table 3.
The correlation between the density of DNMs and various genomic variables at the 1MB scale.
Table 4.
The standardised regression coefficients from a stepwise multiple regression with forward variable selection.
Table 5.
Testing for an effect of paternal age.
Fig 6.
The quality of human-chimp alignments.
The correlation between the divergence from human to chimpanzee and the density of DNMs in humans is plotted against the number of aligned sites per 1MB window for three sets of alignments: UCSC pairwise alignments (PW, blue), UCSC multi-way alignments (MZ, orange) and EPO multi-species alignments (EPO, green).
Fig 7.
Testing why divergence is correlated to recombination rate.
The slope (and SE) between normalised DNM density and normalised sex-averaged recombination rate (RR) (Wong—blue, Francioli–orange, Jonsson–green), normalised substitution density and RR (yellow) and normalised SNP density and RR (light blue). In each case the values were normalised by dividing by the mean.
Table 6.
Testing the difference in slopes.
Fig 8.
The decrease in the correlation between DNM density and divergence with increasing evolutionary divergence.
The graph shows the correlation between DNM density and the substitution rate for different phylogenetic branches (with 95% confidence intervals). Francioli-1, Wong-1 and Jonsson-1 are the correlations involving the divergence along the human lineages from the comparison of human-chimp-orangutan (HCO), human-orangutan-macaque (HOM) and human-macaque-marmoset (HMM). Francioli-2, Wong-2 and Jonsson-2 involve the divergences along the chimpanzee, orangutan and macaque lineages.
Fig 9.
The relationship between DNM density, or divergence, and GC content.
The relationship was estimated from a regression of DNM density or human-chimp divergence, against GC-content and the square of the GC-content at the 1MB scale. Blue–Francioli, Light Orange–Wong, Green–Jonsson and Dark Orange–W<>W and S<>S substitutions between human and chimpanzee.
Table 7.
Correlation between the GC-corrected density of DNMs and various genomic variables.