Figure 2. The distribution of genes classed according to the four promoter types (A) CpG-rich AP, (B) CpG-poor AP, (C) CpG-rich SP and (D) CpG-poor SP.

The standard error for the percentages are calculated using the equation (p*(100−p)/n)^0.5 where p is the percentage of genes that belong to a given promoter type and n is the total number of genes classified. The standard errors for the percentages are plotted as error bars. See Fig S1.