Appendix A — Glossary

ATAC-seq. A method that finds open chromatin: the Tn5 enzyme cuts and tags DNA where it is not packed by nucleosomes (Chapter 1).

BAM. A file of reads aligned to a genome, in compressed form (Chapter 3).

bigWig. A compact file of signal along the genome, for viewing in a genome browser (Chapter 8).

Blacklist. A list of genome regions that collect reads in almost every experiment, for technical reasons. Reads and peaks there are removed (Chapter 3).

Broad mark. A histone mark that covers long stretches of DNA, such as H3K27me3. Compare narrow mark (Chapter 1).

ChIP-seq. A method that finds where a protein or histone mark sits on DNA, using an antibody to pull down the bound DNA (Chapter 1).

Consensus regions. The one set of regions in which reads of all samples are counted for a differential analysis (Chapter 9).

Control. A sample that shows background: an input (DNA without antibody pull-down) for ChIP-seq, or an IgG sample (a non-specific antibody) for CUT&RUN and CUT&Tag (Chapter 1).

CUT&RUN, CUT&Tag. Methods that cut DNA next to an antibody-bound protein inside the cell nucleus. They need fewer cells and reads than ChIP-seq (Chapter 1).

Differential analysis. A test for regions whose signal differs between conditions, taking the variation between replicates into account (Chapter 9).

Dispersion. How much a region’s read count varies between replicates, beyond chance (Chapter 9).

DMR. Differentially methylated region: a region whose DNA methylation differs between conditions (Chapter 17).

Duplicate reads. Reads that start at the same position, usually copies of one DNA molecule made by PCR (Chapter 6).

Enhancer. A DNA region that raises the activity of a gene, often from far away (Chapter 13).

FDR (false discovery rate). The expected share of false results in a list of results. An FDR of 0.1 means that up to about one in ten listed regions is expected to be false (Chapter 9).

Fold change (log₂). How much the signal changed, on a base-2 log scale: 1 means twice as much, −1 half as much (Chapter 9).

FRiP. Fraction of reads in peaks: the share of a library’s reads that fall inside its peaks (Chapter 6).

GO (Gene Ontology). A database of terms that describe what genes do (Chapter 11).

H3K27ac. Acetylation of lysine 27 on histone H3; marks active enhancers and promoters (Chapter 1).

IDR (irreproducible discovery rate). A measure of how consistently peaks rank in two replicates (Chapter 7).

Library complexity. How many different DNA molecules a library contains; measured by NRF, PBC1 and PBC2 (Chapter 6).

MA plot. A plot of each region’s fold change against its mean count (Chapter 9).

MAPQ (mapping quality). A score for how sure the aligner is that a read belongs where it was placed (Chapter 5).

Median-of-ratios (RLE). DESeq2’s default way to compute size factors (Chapter 9).

Motif. A short DNA pattern that a transcription factor prefers to bind (Chapter 12).

Narrow mark. A histone mark with sharp peaks, such as H3K4me3 or H3K27ac (Chapter 1).

Normalisation. Correcting each sample’s counts for how deeply it was sequenced, so that samples can be compared (Chapter 9).

NSC and RSC. Measures of strand cross-correlation that show enrichment in single-end ChIP-seq (Chapter 6).

PCA (principal component analysis). A plot that places samples by how similar their counts are (Chapter 9).

Peak. A region with more reads than expected by chance (Chapter 7).

Pioneer factor. A transcription factor that can bind DNA wrapped around nucleosomes, and help to open it (Chapter 16).

Promoter. The DNA around a gene’s start site, where transcription begins. EpiNexus counts ±3 kb from the start site (Chapter 11).

Replicate. A repeat of a sample. Biological replicates come from different animals, cultures or patients; technical replicates repeat the measurement of the same material (Chapter 2).

RRBS, WGBS. Bisulfite sequencing of a reduced, CpG-rich part of the genome (RRBS), or of the whole genome (WGBS), to measure DNA methylation (Chapter 1).

Shrinkage. Pulling noisy fold changes, mostly of regions with few reads, towards zero (Chapter 9).

Size factor. A number for how deeply a sample was sequenced; counts are divided by it (Chapter 9).

Spike-in. DNA or chromatin from another species, added in a fixed amount to each sample to help normalisation (Chapter 10).

Summit. The point of highest signal within a peak (Chapter 7).

Super-enhancer. A cluster of enhancers with far more signal than typical enhancers, often near genes that define the cell type (Chapter 13).

TMM (trimmed mean of M-values). A way to compute size factors that leaves out the most changed and the most extreme regions (Chapter 9).

TSS (transcription start site). The point where transcription of a gene starts.

TSS enrichment. A measure of ATAC-seq quality: the signal at gene start sites compared with the signal around them (Chapter 6).

Volcano plot. A plot of each region’s fold change against its significance (Chapter 9).