Extended Data Fig. 5: Relationship between Pol III occupancy at tRNA genes with chromatin status and nearby gene activity.
From: Selective gene expression maintains human tRNA anticodon pools during differentiation
a, Heatmaps showing ChIP-Seq signal (RPC1, H3K4me3, H3K27me3, H3K9me3) and nucleosome-free regions (NFR) from ATAC–Seq around tRNA gene start sites (±1 kbp) for single replicates of NPC and CM. Normalized signal, accounting for estimated library sizes generated from counts over extended tRNA features (±125 bp) is scaled to reads-per-million (rpm). tRNA genes are separated into housekeeping, repressed and inactive based on significant peaks in RPC1 ChIP-Seq data (FDR-adjusted P ≤ 0.05), and sorted in descending order based on mean value per region. b, Correlation of mean ATAC–Seq NFR reads aligned to extended tRNA features (±125 bp) to tRNA abundance per deconvoluted unique transcript (n = 373) estimated by mim-tRNAseq (n = 2 biological replicates for each cell type; Pearson’s correlation coefficients). Both metrics are scaled to proportions of total tRNA-mapped reads for each dataset and method. Solid blue lines: linear regression model; shaded grey: 95% confidence interval. c, Violin plots of CpG methylation proportions separated by tRNA activity (centre line: median) from ENCODE. P values are from Wilcoxon tests. d, Boxplot showing the distribution of mean RPC1 ChIP-Seq reads aligned to extended tRNA features (±125 bp) as a function of the distance between tRNA genes in different activity classes to their nearest neighbouring coding gene (centre line and label: median; box limits: upper and lower quartiles; whiskers: 1.5 × interquartile range). Data are separated by activity class of tRNA, and whether the neighbouring coding gene is predicted to be active by presence of upstream H3K4me3 and ATAC–Seq NFR peaks (n = 1 for H3K4me3 in NPC, n = 2 biological replicates for all others). Source numerical data are provided.
