The cellranger aggr
pipeline outputs summary.json
which contains metrics relating to the aggregated datasets. Note: square brackets denote a variable that depends on the pipeline input, e.g. [library_id]_frac_reads_kept
means that if your aggregation contains two libraries with IDs sample123
and sample456
, there will be two output metrics sample123_frac_reads_kept
and sample456_frac_reads_kept
.
Metric | Description |
---|---|
filtered_bcs_transcriptome_union | The estimated number of barcodes associated with cell-containing partitions, summed across all input libraries. |
total_reads | Total number of sequenced reads, summed across all input libraries. |
multi_transcriptome_total_raw_reads_per_filtered_bc | total_reads divided by filtered_bcs_transcriptome_union . |
[library_id]_pre_normalization_raw_reads_per_filtered_bc | The mean total reads per cell prior to depth normalization, for the library denoted by library_id . |
[library_id]_pre_normalization_cmb_reads_per_filtered_bc | The mean confidently mapped and barcoded (CMB) reads per cell prior to depth normalization, for the library denoted by library_id. |
[library_id]_frac_reads_kept | The fraction of reads that were retained after depth normalization for the library denoted by library_id . |
lowest_frac_reads_kept | The lowest fraction of reads retained, corresponding to the library which lost the most reads during normalization. A low value may indicate a large disparity in the initial depth of the input libraries. |