This file just computes the mean over all quality controls statistics described
below over barcodes that are defined with isCell = TRUE. The idea is that this
should represent the overall spectra of the experiment.
DropBarcode: assigned per droplet (after computational pairing) using the previously described formulation.
Exp: shorthand experiment ID for convenience.
FRIP: computed as the fraction of fragments overlapping chromatin accessibility peaks for a given peak annotation (variable dataset to dataset).
nBarcodesDroplet: The number of bead barcodes combined to form the specific droplet barcode.
filteredQ30: The number of reads passing a quality alignment filter of 30.
isDoublet: A boolean from nBarcodesDroplet > 1.
isCell: defined computationally; presently using uniqueNuclearFrags > 1000 and FRIP > 0.5
tssPproportion: proportion of fragments overlapping a TSS +/- 2kb for a given reference set of TSS genes (all Ref-Seq).
meanInsertSize: the mean of the insert sizes for each properly paired read per cell.
medianInsertSize: same but the median.
uniqueNuclearFrags: number of de-duplicated fragments mapping to the nuclear genome.
totalNuclearFrags: number of fragments (including duplicates) mapping to the nuclear genome.
totalMitoFrags: number of fragments uniquely aligning to the mitochondrial genome.
duplicateProportion: defined as (totalNuclearFrags - uniqueNuclearFrags)/ totalNuclearFrags
librarySize: estimated total number of unique reads that can be gathered per cell using the Lander-Waterman computation, a specific case of the "Coupon Collector" problem in statistics.