1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170
|
% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/perCellQCMetrics.R
\name{perCellQCMetrics}
\alias{perCellQCMetrics}
\alias{perCellQCMetrics,ANY-method}
\alias{perCellQCMetrics,SummarizedExperiment-method}
\alias{perCellQCMetrics,SingleCellExperiment-method}
\title{Per-cell quality control metrics}
\usage{
perCellQCMetrics(x, ...)
\S4method{perCellQCMetrics}{ANY}(
x,
subsets = NULL,
percent.top = integer(0),
threshold = 0,
BPPARAM = SerialParam(),
flatten = TRUE,
percent_top = NULL,
detection_limit = NULL
)
\S4method{perCellQCMetrics}{SummarizedExperiment}(x, ..., assay.type = "counts", exprs_values = NULL)
\S4method{perCellQCMetrics}{SingleCellExperiment}(
x,
subsets = NULL,
percent.top = integer(0),
...,
flatten = TRUE,
assay.type = "counts",
use.altexps = TRUE,
percent_top = NULL,
exprs_values = NULL,
use_altexps = NULL
)
}
\arguments{
\item{x}{A numeric matrix of counts with cells in columns and features in rows.
Alternatively, a \linkS4class{SummarizedExperiment} or \linkS4class{SingleCellExperiment} object containing such a matrix.}
\item{...}{For the generic, further arguments to pass to specific methods.
For the SummarizedExperiment and SingleCellExperiment methods, further arguments to pass to the ANY method.}
\item{subsets}{A named list containing one or more vectors
(a character vector of feature names, a logical vector, or a numeric vector of indices),
used to identify interesting feature subsets such as ERCC spike-in transcripts or mitochondrial genes.}
\item{percent.top}{An integer vector specifying the size(s) of the top set of high-abundance genes.
Used to compute the percentage of library size occupied by the most highly expressed genes in each cell.}
\item{threshold}{A numeric scalar specifying the threshold above which a gene is considered to be detected.}
\item{BPPARAM}{A \linkS4class{BiocParallelParam} object specifying how parallelization should be performed.}
\item{flatten}{Logical scalar indicating whether the nested \linkS4class{DataFrame}s in the output should be flattened.}
\item{percent_top, detection_limit, exprs_values, use_altexps}{Soft deprecated equivalents to the arguments described above.}
\item{assay.type}{A string or integer scalar indicating which \code{assays} in the \code{x} contains the count matrix.}
\item{use.altexps}{Logical scalar indicating whether QC statistics should be computed for alternative Experiments in \code{x}.
If \code{TRUE}, statistics are computed for all alternative experiments.
Alternatively, an integer or character vector specifying the alternative Experiments to use to compute QC statistics.
Alternatively \code{NULL}, in which case alternative experiments are not used.}
}
\value{
A \linkS4class{DataFrame} of QC statistics where each row corresponds to a column in \code{x}.
This contains the following fields:
\itemize{
\item \code{sum}: numeric, the sum of counts for each cell.
\item \code{detected}: numeric, the number of observations above \code{threshold}.
}
If \code{flatten=FALSE}, the DataFrame will contain the additional columns:
\itemize{
\item \code{percent.top}: numeric matrix, the percentage of counts assigned to the top most highly expressed genes.
Each column of the matrix corresponds to an entry of \code{percent.top}, sorted in increasing order.
\item \code{subsets}: A nested DataFrame containing statistics for each subset, see Details.
\item \code{altexps}: A nested DataFrame containing statistics for each alternative experiment, see Details.
This is only returned for the SingleCellExperiment method.
\item \code{total}: numeric, the total sum of counts for each cell across main and alternative Experiments.
This is only returned for the SingleCellExperiment method.
}
If \code{flatten=TRUE}, nested matrices and DataFrames are flattened to remove the hierarchical structure from the output DataFrame.
}
\description{
Compute per-cell quality control metrics for a count matrix or a \linkS4class{SingleCellExperiment}.
}
\details{
This function calculates useful QC metrics for identification and removal of potentially problematic cells.
Obvious per-cell metrics are the sum of counts (i.e., the library size) and the number of detected features.
The percentage of counts in the top features also provides a measure of library complexity.
If \code{subsets} is specified, these statistics are also computed for each subset of features.
This is useful for investigating gene sets of interest, e.g., mitochondrial genes, Y chromosome genes.
These statistics are stored as nested \linkS4class{DataFrame}s in the \code{subsets} field of the output.
For example, if the input \code{subsets} contained \code{"Mito"} and \code{"Sex"}, the output would look like:
\preformatted{ output
|-- sum
|-- detected
|-- percent.top
+-- subsets
|-- Mito
| |-- sum
| |-- detected
| +-- percent
+-- Sex
|-- sum
|-- detected
+-- percent
}
Here, the \code{percent} field contains the percentage of each cell's count sum assigned to each subset.
If \code{use.altexps=TRUE}, the same statistics are computed for each alternative experiment in \code{x}.
This can also be an integer or character vector specifying the alternative Experiments to use.
These statistics are also stored as nested \linkS4class{DataFrame}s, this time in the \code{altexps} field of the output.
For example, if \code{x} contained the alternative Experiments \code{"Spike"} and \code{"Ab"}, the output would look like:
\preformatted{ output
|-- sum
|-- detected
|-- percent.top
+-- altexps
| |-- Spike
| | |-- sum
| | |-- detected
| | +-- percent.total
| +-- Ab
| |-- sum
| |-- detected
| +-- percent.total
+-- total
}
The \code{total} field contains the total sum of counts for each cell across the main and alternative Experiments.
The \code{percent} field contains the percentage of the total count in each alternative Experiment for each cell.
If \code{flatten=TRUE}, the nested DataFrames are flattened by concatenating the column names with underscores.
This means that, say, the \code{subsets$Mito$sum} nested field becomes the top-level \code{subsets_Mito_sum} field.
A flattened structure is more convenient for end-users performing interactive analyses,
but less convenient for programmatic access as artificial construction of strings is required.
}
\examples{
example_sce <- mockSCE()
stats <- perCellQCMetrics(example_sce)
stats
# With subsets.
stats2 <- perCellQCMetrics(example_sce, subsets=list(Mito=1:10),
flatten=FALSE)
stats2$subsets
# With alternative Experiments.
pretend.spike <- ifelse(seq_len(nrow(example_sce)) < 10, "Spike", "Gene")
alt_sce <- splitAltExps(example_sce, pretend.spike)
stats3 <- perCellQCMetrics(alt_sce, flatten=FALSE)
stats3$altexps
}
\seealso{
\code{\link{addPerCellQC}}, to add the QC metrics to the column metadata.
}
\author{
Aaron Lun
}
|