Skip to main content
Fig. 1 | Genome Biology

Fig. 1

From: TADsplimer reveals splits and mergers of topologically associating domains for epigenetic regulation of transcription

Fig. 1

Develop the TADsplimer algorithm to detect TAD splits and mergers with high accuracy. a–d Heatmaps showing the chromatin interactions in a fibroblast TAD that was split in HUVEC (a, b) and in a HUVEC TAD that was split in fibroblast (c, d). In each heatmap, the top right triangle area indicates data for the fibroblast IMR90, and the bottom left triangle area indicates data for HUVEC. HUVEC data generated by the same lab for 3 donors was indicated by 3 heatmaps in a and c. HUVEC data generated by an additional lab was indicated in b and d. All heatmaps in a and b indicate data from the same genomic region, whereas all heatmaps in c and d indicate data from another genomic region. The blue circle indicates chromatin loops that were not disrupted by the TAD splits. Color scales for each heatmap were indicated in the top right and bottom left corners. e Cartoons showing steps I to IV for TAD identification in TADsplimer. f Cartoons showing the two steps to define TAD split or merger in one sample relative to another sample in TADsplimer. g ROC curve showing the performance of four alternative methods in TADsplimer for scoring TAD splits. h ROC curve showing the influence of five TAD identification methods on the detection of TAD splits. i Heatmaps showing the simulated frequency of chromatin interaction at a sequencing depth of 400 million (top) or 25 million (bottom) reads. j ROC curve distance to top left corner is plotted against Hi-C sequencing depth to show the performance of the four alternative methods in TADsplimer for scoring TAD splits. k ROC curve distance to top left corner is plotted against Hi-C sequencing depth to show the influence of the five TAD identification methods on detection of TAD splits

Back to article page