Figure 1.

Lineage tree and alternate allele frequencies. (A) The samples in this study by type (rows) and patient (columns). (B) Model of neoplastic progression on the basis of organismal tissue and cell lineage. For simplicity, only one possible scenario of the progression from normal to neoplasia to carcinoma is shown. Mutations that arise in ancestors are propagated through subsequent divisions to all descendants. Depending on the ancestors in which they arise, they will be found in one or more samples of the patient, with varying prevalence. For example, mutations that arise in the B branches will be found in all cells of the neoplasia and of the carcinoma; in contrast, mutations that arise on the C branch will be present only in a subset of the neoplasia cells and mark the neoplastic subpopulation from which the carcinoma arose. Mutations that arise on the F branch mark a clonal expansion within the neoplasia, after the last common ancestor with the carcinoma. Note that if there are no mutations found that define branches B and C, it is not possible to infer a specific relationship of the carcinoma with the neoplasia. (NS) Not sampled. In the expanded box are alternate allele frequency comparisons relevant to neoplasias and carcinomas. The two starred comparisons require independent estimates of the proportion of normal cells in each sample, as they compare AAFs across different samples. All other comparisons are either within samples, or the AAF is zero, thus requiring no independent estimate of the proportion of normal cells in the sample. (C–F) Alternate allele frequencies as a function of the class and sample for each patient with phylogenetically informative SNV-sharing classes. The number of SNVs in each class and the branch in the lineage tree of A are listed below each plot. For Patient 1, the only phylogenetically informative class was where the IDC shared SNVs with ENA. For the other patients, the AAFs of informative classes are grouped together and the mutation pattern for each class is represented by a series of zeros and ones directly above the sample labels (a “1” indicates that the SNVs were present in the corresponding sample and a “0” indicates that they were not). (EN) Early neoplasia; (EN_cl) early neoplasia contralateral; (ENA) early neoplasia with atypia. Subscript in lineage-tree branch of patient 6 denotes whether the neoplasia in the lineage tree is this patient's EN or ENA, and whether the carcinoma is DCIS or IDC.