Version Compatibility

Reference examples tested with: FlowSOM 2.10+, scanpy 1.10+

Before using code patterns, verify installed versions match. If versions differ:

• R: packageVersion('<pkg>') then ?function_name to verify parameters

If code throws ImportError, AttributeError, or TypeError, introspect the installed package and adapt the example to match the actual API rather than retrying.

Clustering and Phenotyping

"Cluster my cytometry data to find cell types" → Discover cell populations in high-dimensional flow/mass cytometry data using unsupervised clustering without predefined gates.

• R: FlowSOM::FlowSOM() for self-organizing map clustering
• R: CATALYST::cluster() with Phenograph or FlowSOM

FlowSOM Clustering

Goal: Cluster cytometry events into cell populations using self-organizing maps.

Approach: Build a FlowSOM grid on marker channels, then extract metacluster assignments per cell.

library(FlowSOM)

# Prepare data
expr <- exprs(fcs)
marker_cols <- grep('CD|HLA', colnames(fcs), value = TRUE)

# Build SOM
fsom <- FlowSOM(fcs,
                colsToUse = marker_cols,
                xdim = 10, ydim = 10,
                nClus = 20,
                seed = 42)

# Get cluster assignments
clusters <- GetMetaclusters(fsom)

# Add to flowFrame
exprs(fcs) <- cbind(exprs(fcs), cluster = clusters)

CATALYST Workflow (Full Pipeline)

Goal: Run the complete CATALYST clustering pipeline from flowSet to annotated cell populations.

Approach: Convert flowSet to SingleCellExperiment with prepData, then cluster on type markers with FlowSOM via CATALYST.

library(CATALYST)
library(SingleCellExperiment)

# Create SCE from flowSet
sce <- prepData(fs, panel, md, transform = TRUE, cofactor = 5)

# Clustering
sce <- cluster(sce,
               features = 'type',  # Use 'type' markers from panel
               xdim = 10, ydim = 10,
               maxK = 20,
               seed = 42)

# View cluster assignments
table(cluster_ids(sce, 'meta20'))