Fig. 4

Classifying Structural Variants into Mutational Vectors. a An example of a bone cancer sample from PCAWG with a highly rearranged genome consisting of both clustered and non-clustered structural variants (SVs) is shown as a Circos plot representation. b Zooming into SVs specifically found on chromosome 12 in the bone cancer sample. SVs are shown as a linear representation (top) and as a rainfall plot (bottom). The rainfall plot depicts all breakpoints on chromosome 12 according to their genomic coordinate (x-axis) and the log₁₀ inter-mutational distance (y-axis), which is the distance to the breakpoint immediately preceding it. The tendency of breakpoints to cluster in a specific genomic region on chromosome 12 due to a chromothripsis event is evident in all representations. c Zooming into SVs specifically found on chromosome 8 in the bone cancer sample. SVs are shown as a linear representation (top) and as a rainfall plot (bottom). The rainfall plot depicts all breakpoints on chromosome 8 according to their genomic coordinate (x-axis) and the log₁₀ inter-mutational distance (y-axis), which is the distance to the breakpoint immediately preceding it. There are no clustered SVs on chromosome 8 as, per the SV classification schema, clustering requires a minimum of 10 breakpoints in a segment of a chromosome. d The SV classification schema is applied to the SVs found on chromosome 12 in the bone cancer sample. SVs are classified by the event type (denoted by color) and are binned according to the size of the event (0 – 10 kb, 10 kb – 100 kb, 100 kb – 1 Mb, 1 Mb – 10 Mb, and > 10 Mb). e The SV classification schema is applied to the SVs found on chromosome 8 in the bone cancer sample. SVs are classified by the event type (denoted by color) and are binned according to the size of the event (0 – 10 kb, 10 kb – 100 kb, 100 kb – 1 Mb, 1 Mb – 10 Mb, and > 10 Mb)