Fig. 2

Fitting the statistical helix model to contact frequencies quantified in mESC. Quantitative 3C data were obtained from wild-type mouse ESC in five gene-rich TADs (a), two gene-poor TADs (b) and one gene-desert TAD (c) (see genomic maps in Additional file 1). For each type of TAD, data obtained from all the anchor primers used for each locus (Additional file 7) were compiled in a single graph (each locus is represented by a specific color). Error bars are standard error of the mean of three independent quantitative 3C assays each quantified at least in triplicate. Dashed lines delimit supranucleosomal domains that encompass separation distances where contact frequencies are alternatively lower and higher (see Methods). The graphs show the best fit analyses obtained with the unconstrained chromatin model [eqs. 1 and 2] (black curves) or the statistical helix model [eqs. 1 and 3] (red curves). Correlation coefficients (R ²) are indicated on the graphs. Best fit parameters, and the genomic distance contained within one statistical helix turn (Sh in kb), are given in the upper part of Table 1. For each supranucleosomal domains, the mean contact frequencies and the number (n) of experimental points are indicated on the graphs. p-values (Mann–Whitney U-test) account for the significance of the differences observed between the experimental means of two adjacent domains (double asterisks indicate a p-value < 0.05 and > 0.01 and triple asterisks a p-value < 0.01)