Closed M4artin closed 1 year ago
There are currently no plans to support this. But it's fairly simple to run the clustering and use FlowSOM
-native functions (as shown here) to get the MST and visualize the SOM. Minimal example below:
# load dependencies
suppressPackageStartupMessages({
library(FlowSOM)
library(flowCore)
library(CATALYST)
})
# setup example data
data(PBMC_fs, PBMC_panel, PBMC_md)
sce <- prepData(PBMC_fs, PBMC_panel, PBMC_md)
# specify grid & markers for clustering
xdim <- ydim <- 10
features <- type_markers(sce)
# construct 'flowFrame'
y <- assay(sce, "exprs")
ff <- flowFrame(t(y))
# get SOM & MST
som <- FlowSOM(ff, colsToUse = features,
xdim = xdim, ydim = ydim, nClus = 20, silent = TRUE,
compensate = FALSE, transform = FALSE, scale = FALSE)
mst <- BuildMST(som)
# visualize with 'PlotStars'
PlotStars(som, backgroundValues = som$metaclustering)
PlotStars(mst)
Hi Helena,
First of all, thank you for a very nice package. I was wondering if there are there any plans for a simple wrapper to produce MST-plots in this package? Or is there a way I can extract it myself along with my own meta cluster annotation (apologies if it's something very basic, I'm rather new to the SCE format).
Kind regards, Martin