Fig. 3

Replication origins identified in a fragment of chromosome 1. A Individual context-dependent somatic mutations used in the ORI detection algorithm (C/A reference allele from cluster A and G/T from cluster B, as shown in Fig. 2), each row represents one individual patient; B mutations of reverse complementary type to those shown on panel A (G/T reference allele from cluster A and C/A from cluster B as shown on Fig. 2); C the consensus PMA score calculated for the combined samples (blue line) and individual sample PMA score (gray), red lines mark the peak positions which represent eight replication origins identified by mORI numbered 1–8; D replication origins identified by other NGS-based methods (see Supp Table 1), in other samples from various tissues; each row represents one sample; except the Akerman SNS-seq track where the top one corresponds to core and bottom to stochastic origins. E RFD profile (blue dots) for two OK-seq samples from [31], along with identified ORI positions (red rectangles); F exons of known genes; G GC content calculated in 1-kb windows; H nucleotide compositional skew profile (gray dots) and replication origin positions from [21]; I replication time obtained for individual ENCODE samples (gray line) and an average for all samples (purple line); J sequence conservation score (phastCons100way); K ENCODE histone marks; L CpG islands (UCSC data); M isochore positions from [32], lowest GC–L1,L2,H1,H2,H3–highest GC; N DNAse hypersensitivity sites obtained for K562 cell line (ENCODE data)