computeSpillmat: Compute spillover matrix

In HelenaLC/CATALYST: Cytometry dATa anALYSis Tools

Compute spillover matrix

Description

Computes a spillover matrix from priorly identified single-positive populations.

Usage

computeSpillmat(
  x,
  assay = "counts",
  interactions = c("default", "all"),
  method = c("default", "classic"),
  trim = 0.5,
  th = 1e-05
)

Arguments

x	a SingleCellExperiment.
assay	character string specifying which assay to use; should be one of assayNames(x) and correspond to count-like data, as linearity assumptions underlying spillover estimation won't hold otherwise.
interactions	"default" or "all". Specifies which interactions spillover should be estimated for. The default exclusively takes into consideration interactions that are sensible from a chemical and physical point of view (see below for more details).
method	"default" or "classic". Specifies the function to be used for spillover estimation (see below for details).
trim	numeric. Specifies the trim value used for estimation of spill values. Note that trim = 0.5 is equivalent to using medians.
th	single non-negative numeric. Specifies the threshold value below which spill estimates will be set to 0.

Details

The default method estimates the spillover as the median ratio between the unstained spillover receiving and the stained spillover emitting channel in the corresponding single stained populations.

method = "classic" will compute the slope of a line through the medians (or trimmed means) of stained and unstained populations. The medians (or trimmed means) computed from events that are i) negative in the respective channels; and, ii) not assigned to interacting channels; and, iii) not unassigned are subtracted as to account for background.

interactions="default" considers only expected interactions, that is, M+/-1 (detection sensitivity), M+16 (oxide formation) and channels measuring metals that are potentially contaminated by isotopic impurites (see reference below and isotope_list).

interaction="all" will estimate spill for all n x n - n interactions, where n denotes the number of single-color controls (= nrow(bc_key(re))).

Value

Returns a square compensation matrix with dimensions and dimension names matching those of the input flowFrame. Spillover is assumed to be linear, and, on the basis of their additive nature, spillover values are computed independently for each interacting pair of channels.

Author(s)

Helena L Crowell [e-mail]

References

Coursey, J.S., Schwab, D.J., Tsai, J.J., Dragoset, R.A. (2015). Atomic weights and isotopic compositions, (available at http://physics.nist.gov/Comp).

Examples

# construct SCE from single-stained control samples
data(ss_exp)
sce <- prepData(ss_exp)

# specify mass channels stained for
bc_ms <- c(139, 141:156, 158:176)

# debarcode single-positive populations
sce <- assignPrelim(sce, bc_ms)
sce <- estCutoffs(sce)
sce <- applyCutoffs(sce)

# estimate & extract spillover matrix 
sce <- computeSpillmat(sce)

library(SingleCellExperiment)
head(metadata(sce)$spillover_matrix)

HelenaLC/CATALYST documentation built on Nov. 30, 2024, 4:04 a.m.