cluster_fcbFlowFrame: Defines populations on barcoded datasets

In bjreisman/DebarcodeR: Automated Demultiplexing for Fluorescent Cell Barcoding

Defines populations on barcoded datasets

Description

This function allows you to calculate the probability of a cell originating from a given population using either gaussian mixture modeling or jenks natural breaks classification

Usage

cluster_fcbFlowFrame(
  fcbFlowFrame,
  channel,
  levels,
  opt = "mixture",
  dist = NULL,
  subsample = 3000,
  trim = 0,
  ret.model = TRUE,
  updateProgress = NULL
)

Arguments

fcbFlowFrame	a fcbFlowFrame object with barcoded flowframe and uptake flowframe post deskewing (at least one barcodes slot filled)
channel	The name (string) of the channel to be clustered
levels	integer, the number of barcoding intensities present in the vector
opt	string, either "mixture" (default) for gaussian mixture modeling, or "fisher" for fisher-jenks natural breaks optimization
dist	string in c("Normal, Skew.normal, Tdist"), passed to mixsmsn
subsample	Integer, number of cells to subsample, defaults to 10,000
trim	numberic between 0, 1; used to trim the upper and lower extremes to exlcude outliers (eg. trim = 0.01 exludes most extreme 1% of data)
ret.model	Option to retain the model for deskewing
updateProgress	used in reactive context (shiny) to return progress information to GUI

Value

a fcbFlowFrame with deskewed barcodes slot and clustering slot with a matrix of probabilities, with ncol = levels, and nrow = legnth(vec). If gaussian mixture modeling is used the probailities correspond to the probability of the cell originaiting that level under the distrubtion specified by the mixture model If jenks natural breaks optimization is used, the probability is estimated empirically based on a histogram

See Also

deskew_fcbFlowFrame

bjreisman/DebarcodeR documentation built on Oct. 22, 2022, 1:50 a.m.