ATAC Cell-by-Peak Data Tables


Files containing putative cells only:

[sample_name]_ATAC_Cell_by_Peak_MEX.zip

Unfiltered file containing all cell indexes with >=1 transposase sites in peaks:

[sample_name]_ATAC_Cell_by_Peak_Unfiltered_MEX.zip

The number of transposase sites from each cell that fall within each peak region is represented in the matrix market exchange (MEX) format. The MEX format is an efficient way to store sparse data, and is a common input format for many single-cell analysis tools. The MEX.zip output files contain three separate gzip compressed files that together represent the number of transposase sites of each peak in each cell. By convention, these files are named:

  • atac-barcodes.tsv: Containing a list of cell indexes (integer between 1 and 3843), one per row.
  • atac-features.tsv: Containing a list of peaks detected, one per row. For improved compatibility, this file contains three columns, the first two of which are a duplicated peak coordinate that follows a format of [chromosome]:[start]-[end], and a third which indicates the features are peaks.
  • atac-matrix.mtx: Containing a three column per row representation of the number of transposase sites in peaks. First column is the 1-based row number from atac-features.tsv (peak). Second column is the 1-based row number from atac-barcodes.tsv (cell). Third column is the number of transposase sites detected for that peak in that cell.

Cell indexes in the atac-barcodes.tsv file are sorted numerically.