Fig. 3: Spatial partitions of subcompartments.
From: Evaluating the role of the nuclear microenvironment in gene function by population-based modeling
a, A representative genome structure showing chromosome folding patterns. Both images show the same structure with different numbers of chromosomes; copies are denoted with (a) and (b). The zoomed inset delineates regions that are primarily occupied by chromatin of the same subcompartment. Subcompartments in each chromatin region are color-coded (A1: pink, A2: yellow, B1: dark blue, B2: green, B3: light blue). b, Procedure to identify spatial partitions of subcompartments: A chromatin interaction network (CIN) is generated from all chromatin regions in a given subcompartment for each structure in the population. Each node in the CIN represents a single chromatin region connected by edges if the two regions are in physical contact in the 3D structure. Nodes are colored by their neighborhood connectivity, ranging from low (blue) to high (red). Highly connected subgraphs were then identified by Markov clustering of CINs (Methods) and visualized in the 3D structure (green dashed circles show examples). The rightmost image illustrates the volume occupied by a spatial partition in a single genome structure. c, Spatial partitions of subcompartments, shown by their occupied volume in the 3D structures. For clarity, only the 50 largest partitions are shown. d, Distributions of the number of partitions per genome structure. e, Distributions of the average size (that is, number of nodes) of subcompartment partitions. In d and e, white circles and black bars show the median value and the interquartile range (IQR: Q3–Q1). Tip points of the violins are minima and maxima. Q1 and Q3 are the lower and upper quartile of the distribution. f, Average fraction of inter-chromosomal edges in spatial partitions. Error bars indicate s.d., and the gray dashed line is the average fraction of all partitions. The number of data points (structures) used in each violin plot and statistics in d, e, and f is 10,000. g, Neighborhood enrichment of chromatin in each subcompartment (Methods). h, A representative structure showing examples of colocalizations of A1–B1 and A2–B3 partitions in the 3D space.
