Fig. 2

Nullomers are observed more frequently than expected by chance in the human genome and various functional components. a Schematic representation of the three nullomer scoring metrics. b φ1 metric score for each nullomer across the genome and the genomic subcompartments for K = 15. c φ1 metric score for each nullomer in the ENCODE cCRE categories for K = 15. d φ2 metric score, log transformed number of occurrences of N nullomers in the genome and various functional categories relative to their occurrences in simulations across K = 10–15 bp. e φ2 metric score log transformed number of occurrences of nullomers in the ENCODE cCRE categories relative to their occurrences in simulations as a function of nullomer length for K = 10–15 bp. For d, e, simulations were performed controlling for trinucleotide content. Error bars represent standard deviation of the number of nullomers obtained from n = 100 simulations of each genomic annotation category and has been log transformed. f φ3 metric score showing the proportion of human nullomers found across 29 other species as a function of nullomer length for K = 11–15 bp. For panels c and e, PLS corresponds to promoter-like signature, pELS to proximal enhancer-like signature, dELS to distal enhancer-like signature, DNase H3K4me3 to DNase and H3K4me3 sites, and CTCF to CTCF-only and CTCF-bound sites