ATAC-seq (Assay for Transposase-Accessible Chromatin using sequencing) is a technique used in molecular biology to assess genome-wide chromatin accessibility.^[1] In 2013, the technique was first described as an alternative advanced method for MNase-seq, FAIRE-Seq and DNase-Seq.^[1] ATAC-seq is a faster analysis of the epigenome than DNase-seq or MNase-seq.^[2][3][4]

ATAC-seq identifies accessible DNA regions by probing open chromatin with hyperactive mutant Tn5 Transposase that inserts sequencing adapters into open regions of the genome.^[2][5] While naturally occurring transposases have a low level of activity, ATAC-seq employs the mutated hyperactive transposase.^[6] In a process called "tagmentation", Tn5 transposase cleaves and tags double-stranded DNA with sequencing adaptors.^[7][8] The tagged DNA fragments are then purified, PCR-amplified, and sequenced using next-generation sequencing.^[8] Sequencing reads can then be used to infer regions of increased accessibility as well as to map regions of transcription factor binding sites and nucleosome positions.^[2] The number of reads for a region correlate with how open that chromatin is, at single nucleotide resolution.^[2] ATAC-seq requires no sonication or phenol-chloroform extraction like FAIRE-seq;^[9] no antibodies like ChIP-seq;^[10] and no sensitive enzymatic digestion like MNase-seq or DNase-seq.^[11] ATAC-seq preparation can be completed in under three hours.^[12]

ATAC-Seq analysis is used to investigate a number of chromatin-accessibility signatures. The most common use is nucleosome mapping experiments,^[3] but it can be applied to mapping transcription factor binding sites,^[13] adapted to map DNA methylation sites,^[14] or combined with sequencing techniques.^[15]

The utility of high-resolution enhancer mapping ranges from studying the evolutionary divergence of enhancer usage (e.g. between chimps and humans) during development^[16] and uncovering a lineage-specific enhancer map used during blood cell differentiation.^[17]

ATAC-Seq has also been applied to defining the genome-wide chromatin accessibility landscape in human cancers,^[18] and revealing an overall decrease in chromatin accessibility in macular degeneration.^[19] Computational footprinting methods can be performed on ATAC-seq to find cell specific binding sites and transcription factors with cell specific activity.^[13]

Modifications to the ATAC-seq protocol have been made to accommodate single-cell analysis. Microfluidics can be used to separate single nuclei and perform ATAC-seq reactions individually.^[12] With this approach, single cells are captured by either a microfluidic device or a liquid deposition system before tagmentation.^[12][20] An alternative technique that does not require single cell isolation is combinatorial cellular indexing.^[21] This technique uses barcoding to measure chromatin accessibility in thousands of individual cells; it can generate epigenomic profiles from 10,000-100,000 cells per experiment.^[22] But combinatorial cellular indexing requires additional, custom-engineered equipment or a large quantity of custom, modified Tn5.^[23] Recently, a pooled barcode method called sci-CAR was developed, allowing joint profiling of chromatin accessibility and gene expression of single cells.^[24]

Computational analysis of scATAC-seq is based on construction of a count matrix with number of reads per open chromatin regions. Open chromatin regions can be defined, for example, by standard peak calling of pseudo bulk ATAC-seq data. Further steps include data reduction with PCA and clustering of cells.^[20] scATAC-seq matrices can be extremely large (hundreds of thousands of regions) and is extremely sparse, i.e. less than 3% of entries are non-zero.^[25] Therefore, imputation of count matrix is another crucial step performed by using various methods such as non-negative matrix factorization. As with bulk ATAC-seq, scATAC-seq allows finding regulators like transcription factors controlling gene expression of cells. This can be achieved by looking at the number of reads around TF motifs^[26] or footprinting analysis.^[25]