Skip to main content
. Author manuscript; available in PMC: 2021 May 20.
Published in final edited form as: Cell Syst. 2020 May 20;10(5):397–407.e5. doi: 10.1016/j.cels.2020.04.004

Figure 4: MATCHA distinguishes multi-way interactions from pairwise interaction cliques.

Figure 4:

(A) Distribution of the occurrence frequencies in the SPRITE data for the Hi-C cliques in each predicted probability group. (B) Overlapping between the Hi-C cliques with different predicted probability scores and the scHi-C cliques. The bar indicates the percentage of the total number of Hi-C cliques and scHi-C cliques in each group. (C) An example where the triplets predicted by MATCHA are not defined as positive samples based on occurrence frequency in SPRITE. The heatmap at the top shows the Hi-C O/E contact matrix. For the triplets track, each line represents a triplet either predicted by MATCHA or from the SPRITE data. Repli-seq signals as well as Hi-C subcompartments and super-enhancer annotations for GM12878 are also shown. Note that the left and middle anchor regions each have 5 genes that are regulated by the same 12 transcription factors, respectively. All 12 transcription factors are confirmed by ChIP-seq data to show that they bind to the upstream of transcriptional start sites (within 3kbp) of the genes from the left and middle anchor regions and also bind to the super-enhancer in the right anchor. A zoom-in view of the related histone modifications and transcription factor ChIP-seq signals are included to show one of the co-regulation examples (bottom panel). See also Figure S1.