Detecting outliers within the SingleCellExperiment object.

Source: R/detectCellOutlier.R

A wrapper function for isOutlier. Identify outliers from numeric vectors stored in the SingleCellExperiment object.

detectCellOutlier(
  inSCE,
  slotName,
  itemName,
  sample = NULL,
  nmads = 3,
  type = "both",
  overwrite = TRUE
)

Arguments

inSCE

A SingleCellExperiment object.

slotName

Desired slot of SingleCellExperiment used for plotting. Possible options: "assays", "colData", "metadata", "reducedDims". Required.

itemName

Desired vector within the slot used for plotting. Required.

sample

A single character specifying a name that can be found in colData(inSCE) to directly use the cell annotation; or a character vector with as many elements as cells to indicates which sample each cell belongs to. Default NULL. decontX will be run on cells from each sample separately.

nmads

Integer. Number of median absolute deviation. Parameter may be adjusted for more lenient or stringent outlier cutoff. Default 3.

type

Character. Type/direction of outlier detection; whether the lower/higher outliers should be detected, or both. Options are "both", "lower", "higher".

overwrite

Boolean. If TRUE, and this function has previously generated an outlier decision on the same itemName, the outlier decision will be overwritten. Default TRUE.

Value

A SingleCellExperiment object with '' added to the colData slot. Additionally, the decontaminated counts will be added as an assay called 'decontXCounts'.

Examples

data(scExample, package = "singleCellTK")
sce <- subsetSCECols(sce, colData = "type != 'EmptyDroplet'")
sce <- runDecontX(sce[,sample(ncol(sce),20)])
#> Thu Mar 31 12:07:23 2022 ... Running 'DecontX'
#> --------------------------------------------------
#> Starting DecontX
#> --------------------------------------------------
#> Thu Mar 31 12:07:24 2022 .. Analyzing all cells
#> Thu Mar 31 12:07:24 2022 .... Generating UMAP and estimating cell types
#> Thu Mar 31 12:07:26 2022 .... Estimating contamination
#> Thu Mar 31 12:07:26 2022 ...... Completed iteration: 10 | converge: 0.01482
#> Thu Mar 31 12:07:26 2022 ...... Completed iteration: 20 | converge: 0.003821
#> Thu Mar 31 12:07:26 2022 ...... Completed iteration: 30 | converge: 0.002048
#> Thu Mar 31 12:07:26 2022 ...... Completed iteration: 40 | converge: 0.00169
#> Thu Mar 31 12:07:26 2022 ...... Completed iteration: 50 | converge: 0.001567
#> Thu Mar 31 12:07:26 2022 ...... Completed iteration: 60 | converge: 0.001433
#> Thu Mar 31 12:07:26 2022 ...... Completed iteration: 70 | converge: 0.001297
#> Thu Mar 31 12:07:26 2022 ...... Completed iteration: 80 | converge: 0.001173
#> Thu Mar 31 12:07:26 2022 ...... Completed iteration: 90 | converge: 0.00106
#> Thu Mar 31 12:07:26 2022 ...... Completed iteration: 96 | converge: 0.0009945
#> Thu Mar 31 12:07:26 2022 .. Calculating final decontaminated matrix
#> --------------------------------------------------
#> Completed DecontX. Total time: 2.290453 secs
#> --------------------------------------------------
sce <- detectCellOutlier(sce, slotName = "colData", sample = sce$sample,
 nmads = 4, itemName = "decontX_contamination", type = "both")