Properties of the N-domains detected in the human genome. (A) Examples of N-domains detected in the chromosome 13. S values are computed in 1-kbp windows (without repeats); (red) + genes; (blue) − genes; (black) intergenic regions; the N-domain borders are indicated by red vertical lines. In abscissa, the window position is in megabase pairs; in ordinate, the skew, S, in percent. (B) Mean S profile of the N-domains. The mean S values are computed along the N-domains of length L ≤ 1.2 Mbp. In abscissa, the region used for analysis extends from the extremity to the center of each domain. In ordinate, the mean skew, S, in percent ± SEM. (C) Mean S profile of the half-domains for L < 0.75 Mbp (red), 0.75 < L < 1.2 Mbp (blue), 1.2 < L < 2 Mbp (purple), and L > 2 Mbp (green). The sequences of the 3′ halves of the domains are reverse-complemented and analyzed together with the 5′ halves. (D) Mean skew profile of + genes located in 5′ half-domains analyzed together with − genes (reverse-complemented) located in 3′ halves (red) and intergenic regions (black) (both larger than 400 kbp, and situated in domains with L > 1 Mbp). In abscissa, the distance ∂ to the 5′ end of genes or intergenic regions. (E) Mean slope of the domains versus their length L; domains are ranked by L values and grouped by bins of 20 domains; in ordinate, the mean (±SEM) of the slopes in percent/megabase pair (orange); the orange hyperbolic curve is obtained by a linear regression fit of −1/slope versus L (Supplemental Fig. S5f). In red, the genes with a length >400 kbp are ranked by length of their domain, and grouped by constant bins; the mean slope is computed for each bin. The same is true for the intergenic regions (>400 kbp) (black). In abscissa, the mean length of the corresponding domains.
