Fig. 5: EZH2 mutation status can change over time in follicular lymphoma with concomitant changes in H3K27me3 distribution.

a H3K27me3 ChIP-seq tracks for follicular lymphoma (FL) samples from individual patients at distinct time points (see Fig. 4b). b Mean density plots of H3K27me3 ChIP-seq reads for FL samples from individual patients at distinct time points, across H3K27me3 peaks identified in the analysis presented in Fig. 4. EZH2-mutant data are color-coded according to VAF as indicated. Note that patient 28, time point #1 is also part of the analysis presented in Fig. 4. VAF, variant allele frequency. c Position of H3K27me3 ChIP-seq samples from 10 individual patients at distinct time points along principal component 1 (PC1) when mapped onto a principal component analysis (PCA) of variance among the full set of H3K27me3 ChIP-seq performed on FL samples, considering only the 5000 most variable H3K27me3-enriched regions, as shown in Fig. 4g. Positions of centroids of EZH2-WT and -mutant subsets along PC1 are indicated. (Full PCA plot shown in Supplementary Fig. 5a.) d Position of RNA-seq samples from 8 individual patients at distinct time points (see Fig. 4b) along principal component 1 (PC1) when mapped onto a PCA of variance among the full set of RNA-seq experiments performed on FL samples, considering only the 211 differentially expressed genes displayed in Fig. 4c. Positions of centroids of EZH2-WT and -mutant subsets along PC1 are indicated. (Full PCA plot shown in Supplementary Fig. 5b.) e Frequency and type of mutations affecting 31 genes among 10 patients from whom multiple samples were analyzed by ChIP-seq. Rows represent genes and columns represent patient samples. For each sample, bars above show the number of genes containing a mutation. For each gene, bars to the right show the number of samples carrying a mutation, and bars to the left show the mean VAF among samples with mutations.