Supplementary Figure 14 : Using CAPRA to dissect combinatorial activities of cCREs

Supplementary Figure 14 | Using CAPRA to dissect combinatorial activities of cCREs

a, Density plots showing the distribution of STARR scores calculated by CAPRA from solo (blue) and double (purple) fragments. b, Barplots showing the pearson correlation coefficient (PCC) for cCRE pairs from the average STARR scores calculated from testing individual cCREs (solo DNA fragments) and the STARR scores calculated from testing multiple elements (double DNA fragments) as the threshold for the minimum number of input DNA fragments increases. Representative scatterplots for a cutoff of one (red box) and ten (green box) are shown in c and d. c, Density scatterplot of STARR scores for pairs of cCREs with a minimum of one DNA input fragment calculated by averaging quantifications from solo fragments or double fragments. d, Density scatterplot of STARR scores for pairs of cCREs with a minimum of ten DNA input fragments calculated by averaging quantifications from solo fragments or double fragments. e, Example of two cCREs (1: EH38E3837011 and 2: EH38E3837012) within the TTPA gene with lower than expected combinatorial effects. These cCRE have high DNase and H3K27ac signals in liver but low signals in K562 cells. f, STARR scores for EH38E3837011 and EH38E3837012 deriving from solo fragments (top) and double fragments (bottom). Scores with p < 0.05 are in bold.