Skip to main content
Figure 1 | Genome Biology

Figure 1

From: Fast and scalable inference of multi-sample cancer lineages

Figure 1

LICHeE method overview.(a) Toy example of five samples (lymph control and four tumor samples S1 to S4) with one germ-line single nucleotide variant (SNV) and four SSNVs, each associated with a binary sample profile. (b) SSNVs are partitioned into groups based on their binary profile (displayed groups contain other SSNVs with varying VAFs). (c) SSNV groups are clustered based on their VAFs. (d) The clusters of each SSNV group are incorporated into an evolutionary constraint network. An edge is placed between a parent node and child node if, for each sample, the VAF of the parent is greater than or equal to that of the child node. For example, this constraint is violated for the green 01110 node and node 00110. (e) The lineage tree is constructed from the constraint network. The leaves of the tree are the individual samples. The horizontal line subdivisions in a sample indicate mixed lineages, separating the different subpopulations of cells in the sample. The colors in each subdivision describe the mutation groups that the cells in this subpopulation have. Consider the orange node 01110, which denotes SSNVs of class 01110. Those SSNVs are found in samples 1, 2, and 3. After an ancestral cell division, the daughter cells’ lineages accumulated SSNVs too (01110, green; 00110) that are now present in their descendent samples or subclones. About 20% of sample 1 are cells that come from the orange and green lineages, and about 40% come from the blue lineage. Samples 1, 2, and 3 grew from two or more subclones, whereas sample 4 only grew from one subclone. SNV, single nucleotide variant; SSNV, somatic single nucleotide variant; VAF, variant allele frequency.

Back to article page