Figure 10

High resolution dendrograms generated from iterated ROA dictionaries provide historical perspective on tumor subclone development. These dendrograms from verified words in FL-128 BCL2 ROA 171 to 204 dictionary depict an inferred developmental relationship between putative tumor subclones, with nodes depicting observed verified word sequences and lines representing mutational differences between words. The relative population fraction is reflected by node circle area and genetic distance from the normal reference sequence by horizontal displacement. Mutations are noted by position and called SNV. (A) Seven clones are found in specimen 128 using BFAST mapping without iteration in which an ambiguous developmental history of the most frequent clone (MFC) is seen along with a single descendant. Mapped coverage is 14,623 reads. (B) Five clones are found in specimen 128 by iterating BFAST to convergence (4 mapping iterations) in which the most frequent clone has two low frequency descendants. Mapped coverage is 15,492 reads. (C) Seven clones are found in specimen 128 by mapping first with BFAST, followed by SHRiMP2, and then iterating BFAST to convergence (BSBn method, 4 mapping iterations) in which a 5.7% clone and its 0.3% descendant containing mutations at two adjacent positions (202 and 203) were revealed by SHRiMP2. Mapped coverage is 15,594 reads.