ATAC Library Metrics Overview

The ATAC pipeline uses SnapATAC2 to generate library-level metrics in CSV format.

Metric	Description
NHash_ID	A unique identifier used to track and reference the specific sample or dataset.
sequenced_reads	The total number of reads generated from the sequencing process, which includes both reads that are mapped and unmapped.
sequenced_read_pairs	The total number of read pairs (two reads per pair) generated from the sequencing process. This is typically half of the total sequenced reads if all reads are paired.
fraction_valid_barcode	The fraction of reads that contain a valid barcode, indicating the proportion of reads that are correctly assigned to a specific cell or sample.
fraction_Q30_bases_in_read_1	The proportion of bases in Read 1 that have a Phred quality score of 30 or higher, indicating high-confidence base calls.
fraction_Q30_bases_in_read_2	The proportion of bases in Read 2 that have a Phred quality score of 30 or higher, indicating high-confidence base calls.
number_of_cells	The estimated number of cells captured and sequenced in the experiment, based on the barcodes identified. Cells are called with SnapATAC2's OrdMag method on fragments overlapping peaks. By default the method estimates the expected cell count from the data; when anchor_cell_calling is true, it uses atac_expected_cells instead. The per-cell metrics below are computed over the same cells as the h5ad output.
mean_raw_read_pairs_per_cell	The average number of raw read pairs associated with each cell, providing an indication of the sequencing depth per cell.
median_high-quality_fragments_per_cell	The median number of high-quality (e.g., confidently mapped) fragments associated with each cell, representing typical fragment quality across cells.
fraction of high-quality fragments in cells	The fraction of high-quality fragments that are associated with identified cells, indicating the proportion of good-quality data that is cell-associated.
fraction_of_transposition_events_in_peaks_in_cells	The fraction of transposition events within identified cells that occur within peaks, which are regions of accessible chromatin.
fraction_duplicates	The fraction of sequenced fragments that are duplicates, which can result from PCR amplification or other factors, indicating the redundancy in the sequencing data.
fraction_confidently_mapped	The fraction of sequenced fragments that are confidently mapped to the reference genome, indicating the proportion of reads that align well to the genome.
fraction_unmapped	The fraction of sequenced fragments that could not be mapped to the reference genome, which can indicate sequencing errors, contamination, or regions not covered by the reference.
fraction_nonnuclear	The fraction of sequenced fragments that are mapped to non-nuclear (e.g., mitochondrial or other organellar) DNA, providing insight into contamination or organellar activity.
fraction_fragment_in_nucleosome_free_region	The fraction of sequenced fragments that map to nucleosome-free regions, which are indicative of accessible chromatin.
fraction_fragment_flanking_single_nucleosome	The fraction of sequenced fragments that map to regions flanking single nucleosomes, indicating regions with partial chromatin accessibility.
tss_enrichment_score	A measure of the enrichment of transposition events at transcription start sites (TSS), indicating the accessibility of promoters across the genome.
fraction_of_high-quality_fragments_overlapping_TSS	The fraction of high-quality fragments that overlap transcription start sites (TSS), providing insight into promoter accessibility.
Number_of_peaks	The total number of peaks, or regions of accessible chromatin, identified in the dataset, representing potential regulatory elements.
fraction_of_genome_in_peaks	The fraction of the genome that is covered by identified peaks, indicating the extent of chromatin accessibility across the genome.
fraction_of_high-quality_fragments_overlapping_peaks	The fraction of high-quality fragments that overlap with identified peaks, providing an indication of the efficiency of the assay in capturing accessible regions.
atac_percent_target	Percent of cells recovered; value is calculated as estimated_cells/expected_cells. The value can exceed 100% when more cells are called than atac_expected_cells, which is common when anchor_cell_calling is true, because anchoring does not cap the number of cells called. A value well above 100% suggests atac_expected_cells is lower than the number of nuclei in the library, or that some called barcodes are not real cells.