tof_preprocess: Preprocess raw high-dimensional cytometry data.

In keyes-timothy/tidytof: Analyze High-dimensional Cytometry Data Using Tidy Data Principles

Preprocess raw high-dimensional cytometry data.

Description

This function transforms a 'tof_tbl' of raw ion counts, reads, or fluorescence intensity units directly measured on a cytometer using a user-provided function. It can be used to perform standard pre-processing steps (i.e. arcsinh transformation) before cytometry data analysis.

Usage

tof_preprocess(
  tof_tibble = NULL,
  channel_cols = where(tof_is_numeric),
  undo_noise = FALSE,
  transform_fun = function(x) asinh(x/5)
)

Arguments

tof_tibble	A 'tof_tbl' or a 'tibble'.
channel_cols	Unquoted column names representing columns that contain single-cell protein measurements. Supports tidyselect helpers. If nothing is specified, the default is to transform all numeric columns.
undo_noise	A boolean value indicating whether to remove the uniform noise that Fluidigm software adds to CyTOF measurements for aesthetic and visualization purposes. See this paper. Defaults to FALSE.
transform_fun	A vectorized function to apply to each protein value for variance stabilization. Defaults to asinh transformation (with a co-factor of 5).

Value

A 'tof_tbl' with identical dimensions to the input 'tof_tibble', with all columns specified in channel_cols transformed using 'transform_fun' (with noise removed or not removed depending on 'undo_noise').

See Also

[tof_postprocess()]

Examples

# read in an example .fcs file from tidytof's internal datasets
input_file <- dir(tidytof_example_data("aml"), full.names = TRUE)[[1]]
tof_tibble <- tof_read_data(input_file)

# preprocess all numeric columns with default behavior
# arcsinh transformation with a cofactor of 5
tof_preprocess(tof_tibble)

# preprocess all numeric columns using the log base 10 tranformation
tof_preprocess(tof_tibble, transform_fun = log10)

keyes-timothy/tidytof documentation built on Aug. 28, 2024, 8:37 a.m.