cluster.flow: Use kmeans clustering to detect subpopulations in a flow...

In ctriandafillou/flownalysis: Analysis of Yeast Protein Induction and pH by Flow Cytometry

Description

Use kmeans clustering to detect subpopulations in a flow cytometry experiment: implements methodology from the flowPeaks package by Ge Y. et al: see citation("flowPeaks")

Usage

cluster.flow(flow.df, cluster.on, tol.u = 1, h0.u = 1, h.u = 1.5,
  qc.exp = FALSE, qc.ax1 = NULL, qc.ax2 = NULL)

Arguments

flow.df	the name of the flow experiment in the form of a dataframe (assumed output from merge_flowSet)
cluster.on	the column names to cluster on; currently only clustering on subsets of the data is allowed
tol.u	user-defined tolerance for clusters (goes to flowPeaks function); default is 1 (can go from 0 to 1)
h0.u	user-defined h0 parameter; default is 1
h.u	user-defined h parameter; default is 1.5
qc.exp	optional qc experiment to generate a plot for to visually inspect clusters [currently not working]
qc.ax1	x axis of qc plot (must be a column name in flow.df)
qc.ax2	y axis of qc plot (must be a column name in flow.df)

Value

the input dataframe flow.df with an additional column called 'cluster' that contains the cluster assignment for each observation

ctriandafillou/flownalysis documentation built on Sept. 7, 2020, 1:48 a.m.