An R package to produce full spectrum flow cytometry plots outside the acquisition software.

Author: Christopher Hall, Babraham Institute, UK

Babraham Institute Flow Cytometry Facility

The purpose of this package is to produce full spectrum flow cytometry plots outside the acquisition software. There are three styles designed to roughly mimic the output styles of the Bigfoot, Aurora, and R Viridis. The package takes flowFrames and flowSets as produced using the flowCore package.

• flowSpectrum wiil not work with the Sony SP6800 as it uses a proprietary format that flowCore can't read.

spectralplot(ff)

Load data into R using flowCore either via read.FCS(), read.flowSet(), or read.ncdfFlowSet.

Use the spectralplot() function, choose a style, and choose where to output the resulting image. The default (i.e. just spectralplot(ff)) will produce a viridis plot in the R session.

## install from Github using devtools (or remotes)
devtools::install_github('hally166/flowSpectrum')

## load the package
library(flowSpectrum)

## flowSpectrum has one demo file that can be loaded.  
ff<-read.FCS(system.file("extdata", "PE.fcs", package = "flowSpectrum"))

## create plots using spectralplot(). spectralpolt() will accept a flowFrame or a flowSet
spectralplot(ff, theme='viridis', save=FALSE, bins=512, normalize=FALSE, params=NULL, guessPop=FALSE, unstained=NULL)

## if you want flowSpectrum to select only the positive spectrum use guessPop.  I advice using normalize = TRUE as it only selects 200 events.
ff<-read.FCS("C:/sample1.fcs)
ctrl_ff<-read.FCS("C:/unstained.fcs)
spectralplot(ff, normalize=TRUE, guessPop=TRUE, unstained=ctrl_ff)

The theme options are 'viridis', 'bigfoot', and 'aurora'.

The save options are TRUE and FALSE. TRUE will save a PNG into the working directory. FALSE will output to the R session.

Set then granularity of the plot using bins = [a number]. I use something between 256 and 512.

Normalize = TRUE will produce a normalized spectrum based on the max median intensity.

Specify which parameters to plot (in the order specified) using params = [a character vector of parameter names].

Use guessPop = TRUE and normalize = TRUE to select the positive events from a mixed control. Add a file to unstained.

Defaults are: theme='viridis', save=FALSE, bins=512, normalize=FALSE, params=NULL, guessPop=FALSE, unstained=NULL

To use the params argument you must supply a character vector (not a list). For example, to reorder to the same as the source file for the Aurora params <- grep("-W|-H|Time|SC",ff@parameters@data$name,value = TRUE, invert=TRUE)