Fig. 2

Reproducible selection of genomic subclones in lung metastases. a PIM1-CBRLuc primary tumors and lung metastases were isolated from 3 tumor-bearing mice (labeled either blue, red, or green). BLI was used to isolate distinct lung metastases in each mouse. Lu, lung metastasis. b RNA isolated from MFP tumors and lung metastases was subjected to RNA-Seq. The top lung metastasis-altered pathways (GeneGo MetaCore) are displayed. c Targeted sequencing validation was used to calculate the average change in MAF between lung metastases and MFP tumors. The horizontal dotted line denotes a p-value of <0.05. Mutations significantly enriched or depleted in lung metastases by a change in MAF of at least 0.15 are shown in green or red, respectively. d Non-silent mutations identified by WES and validated by targeted sequencing significantly enriched or depleted (p-value < 0.05, ΔMAF at least 0.15) in lung metastases are shown in the heat map of MAFs calculated from targeted sequencing organized by hierarchical clustering. Predicted amino acid changes are denoted on the right. Mutations predicted to be damaging or potentially damaging to protein function are indicated in red, tolerated or benign mutations are indicated in gray. Mutations detected by RNA-seq are indicated in green, mutant loci covered in RNA-seq without detection of the mutation are indicated in black, and loci without sufficient coverage for assessment of MAF by RNA-seq are indicated in gray. *Mutations falling in COSMIC cancer genes. e Average MAFs of validated mutations significantly enriched in lung metastases compared to MFP tumors are plotted. f Average MAFs of validated mutations significantly depleted in lung metastases compared to MFP tumors are plotted. g The cellular prevalence of mutation clusters as estimated by PyClone are plotted. Line thickness is proportional to the number of mutations within the cluster. The number of mutations comprising each cluster is shown in parentheses. h Subclonal architecture was manually reconstructed based on PyClone results. Each pie chart represents a tumor sample, and the colors in the pie chart represent subclones harboring distinct sets of mutation clusters (see g)