Fig. 5: DNA tracing in Mb-scale regions.
a Overview of a genomic region (Chr7-SHH) targeted with 12 probes (highlighted by coloured circles) across 1.6 Mb. Each probe targets a ~ 10 kb-sized region. Major CTCF sites (HeLa CTCF Chip-seq) predicted to form loops are indicated by the dashed green line. b Consensus representations and single trace examples from HeLa WT or HeLa RAD21-mEGFP-AID, CTCF-mEGFP-AID treated for 2 h with auxin. Data from 3 (WT) or 2 (AID cells) replicates, number of traces per condition are indicated. Green dashed circles indicate <100 nm proximity between probe no. 3 at the TAD boundary and downstream CTCF sites, while orange circle indicate three-way contacts (stacked loops). Grey spheres indicate standard deviation of each position compared to consensus representation. c Number of contacts (<100 nm) per trace for four 1.3-1.8 Mb regions on chr2, 4 and 7 in HeLa WT, ΔRAD21 and ΔCTCF cells. * p < 0.05 for condition vs wild type, tested by Kruskal-Wallis non-parametric 1-way ANOVA followed by Conover’s test with Holm adjustment for multiple comparisons. Violin plots show median values, boxes indicate quartiles and whiskers show data range within 1.5 times IQR. d UMAP with clustering of single traces by a Bayesian GMM (see methods) from the chr7-SHH locus in HeLa WT cells. Median distance reconstructions of exemplary clusters and insets highlighting the internal TAD architecture (3rd – 9th probe positions, dominating loops highlighted with green dashed ring) as indicated by trace/cluster colour. In total n = 2290 traces from 3 independent experiments, n = 141 (grey), n = 174 (red), n = 365 (brown) and n = 273 (purple) traces in the highlighted clusters. See Supplementary Fig. S6f for all cluster medians. Scale bars are 100 nm. Source data are provided as a Source Data file.
