Figure 2

Observed distributions of p and q values. (a) The distribution of p values for the same data as in Figure 1. They are relatively uniformly distributed on the interval (0,1), except for a slight peak close to p = 0, indicating a small fraction of specifically bound intergenic regions (IGRs). (b) Corresponding q values, but with a log scale on the vertical axis. As one descends the ranked list of IGRs the q value rapidly approaches unity. That most IGRs have q close to 1 is expected given that the list of tested IGRs is long, and the number of true targets is generally small.