CytoNorm.train: CytoNorm.train

In saeyslab/CytoNorm: Normalisation of cytometry data measured across multiple batches

CytoNorm.train

Description

CytoNorm learns batch effects, using control samples which should be the the same per batch. The cells are first split into rough groups using the FlowSOM algorithm, after which splines are trained to map the quantiles for each control group to an average distribution Typically, you will use the function CytoNorm.normalize after this function to reverse the batch effects in other files.

Usage

CytoNorm.train(
  files,
  labels,
  channels,
  transformList,
  outputDir = "./tmp",
  FlowSOM.params = list(nCells = 1e+06, xdim = 15, ydim = 15, nClus = 10, scale = FALSE),
  normMethod.train = QuantileNorm.train,
  normParams = list(nQ = 99),
  seed = NULL,
  clean = TRUE,
  plot = FALSE,
  verbose = FALSE,
  recompute = FALSE,
  ...
)

Arguments

files	Full paths to the fcs files or a flowSet of the control samples.
labels	A label for every file, indicating to which batch it belongs, e.g. the plate ID.
channels	Column names of the channels that need to be normalized
transformList	Transformation list to pass to the flowCore transform function
outputDir	Directory to put the temporary files in. Default = "./tmp"
FlowSOM.params	Extra parameters to be passed to the FlowSOM function, such as the number of cells (nCells), the xdim and y dim or the number of clusters (nClus)
normMethod.train	Normalization method to use for each cluster. Default = QuantileNorm.train
normParams	Parameters to pass to the normalization method. Default, assuming QuantileNorm.train: list(nQ = 99)). nQ is the number of quantiles to use, 0.01 to 0.99 by default.
seed	Set a seed for reproducable results.
clean	Whether to remove the temporary files again at the end. Default = TRUE.
plot	If TRUE, plots are saved to the current dir. Default = FALSE.
verbose	If TRUE, progress updates are printed. Default = FALSE.
recompute	If FALSE, will try to reuse previously saved FlowSOM model. If so, a warning message will be printed. Default = FALSE.
...	Additional arguments to pass to read.FCS

Details

Temporary fcs files splitting the data into clusters are written to outputDir and will be removed again by default (depending on clean).

Value

A list containing two elements: the FlowSOM clustering and a list containing all the splines per cluster. This can be used as input for the CytoNorm.normalize function.

See Also

QuantileNorm.train, CytoNorm.normalize, prepareFlowSOM

Examples

dir <- system.file("extdata", package = "CytoNorm")
files <- list.files(dir, pattern = "fcs$")
data <- data.frame(File = files,
                   Path = file.path(dir, files),
                   Type = stringr::str_match(files, "_([12]).fcs")[,2],
                   Batch = stringr::str_match(files, "PTLG[0-9]*")[,1],
                   stringsAsFactors = FALSE)
data$Type <- c("1" = "Train", "2" = "Validation")[data$Type]
train_data <- dplyr::filter(data, Type == "Train")
validation_data <- dplyr::filter(data, Type == "Validation")

ff <- flowCore::read.FCS(data$Path[1])
channels <- grep("Di$", flowCore::colnames(ff), value = TRUE)
transformList <- flowCore::transformList(channels,
                                         cytofTransform)
transformList.reverse <- flowCore::transformList(channels,
                                                 cytofTransform.reverse)

model <- CytoNorm.train(files = train_data$Path,
                        labels = train_data$Batch,
                        channels = channels,
                        transformList = transformList,
                        FlowSOM.params = list(nCells = 10000, #1000000
                                              xdim = 10,
                                              ydim = 10,
                                              nClus = 10,
                                              scale = FALSE),
                        normParams = list(nQ = 99),
                        seed = 1)

CytoNorm.normalize(model = model,
                   files = validation_data$Path,
                   labels = validation_data$Batch,
                   transformList = transformList,
                   transformList.reverse = transformList.reverse)

saeyslab/CytoNorm documentation built on Nov. 2, 2024, 12:39 p.m.