Cytometry quickstart
Step-by-step workflows for FlowSOM clustering and IMC segmentation and spatial analysis.
Retired tools
flow_gate, flow_cluster and flow_differential are retired and can no longer be run. For FCS clustering use flowsom. There is no hosted tool for automated gating or differential abundance; ask Ivy to run those in Python in the chat's code workspace.
You want reproducible populations and spatial biology out of your cytometry data. This guide runs the fastest path through three workflows — FlowSOM clustering, IMC cell segmentation, and IMC spatial analysis — each in a few steps.
Cytometry quickstart
This guide covers three common workflows: FlowSOM clustering, IMC cell segmentation, and IMC spatial analysis.
Configure-manually tables go beyond the implemented tools
The IMC "configure manually" parameter tables below describe the intended interface, not the one imc_segment and imc_spatial currently accept. See the note on the Cytometry overview for what each tool accepts.
Workflow 1: FlowSOM clustering
Upload your FCS file
Upload an FCS file through the chat interface or data panel. Flow cytometry data should be compensated before upload; CyTOF data does not need compensation.
Run clustering
Or configure:
| Parameter | Value |
|---|---|
fcs_key | The uploaded FCS file |
cols_to_use | ["CD3", "CD4", "CD8", "CD19", "CD56", "CD14", "CD16"] |
n_clusters | 20 |
xdim / ydim | 10 / 10 |
Use lineage markers (CD3, CD4, CD8, CD19, CD56, etc.) for population identification.
Review cluster results
flowsom returns the size and fraction of events in each metacluster, the median marker intensities per metacluster, per-event cluster labels as a CSV, and an optional star plot of the SOM grid. Read the high-intensity markers of each metacluster to name it.
Compare groups (optional)
There is no hosted differential-abundance tool. Take the per-event cluster labels from flowsom and ask Ivy to run the group comparison in the chat's code workspace.
Workflow 2: IMC cell segmentation
Upload your IMC image
Upload a multi-channel TIFF or MCD file containing your imaging mass cytometry acquisition.
Run segmentation
Or configure:
| Parameter | Value |
|---|---|
| Method | cellpose |
| Nuclear markers | ["Ir193"] |
| Membrane markers | ["CD45", "Na-K-ATPase"] |
| Min cell area | 50 pixels |
| Max cell area | 5000 pixels |
Review segmentation results
| Field | Example Value |
|---|---|
| Total cells segmented | 3,847 |
| Mean cell area | 312 px |
| Median markers per cell | 35 |
| Oversegmentation rate | 2.1% |
| Undersegmentation rate | 4.3% |
Extract per-cell marker intensities
The segmentation output includes a cell-by-marker matrix with mean intensities for each channel within each cell mask. This matrix is the input for downstream clustering and spatial analysis.
Workflow 3: IMC spatial analysis
Prepare segmented data
Use the output from imc_segment, or provide your own cell data with X/Y coordinates and cell type labels.
Run spatial analysis
Or configure:
| Parameter | Value |
|---|---|
| Analysis type | neighborhood |
| Cell types | ["CD8_T", "Tumor", "Macrophage", "Stroma"] |
| Radius | 30 microns |
| Permutations | 1000 |
Review spatial results
| Cell Type A | Cell Type B | Interaction Score | p-value | Relationship |
|---|---|---|---|---|
| CD8 T | Tumor | 2.34 | 0.001 | Attraction |
| Macrophage | Tumor | 1.87 | 0.008 | Attraction |
| CD8 T | Stroma | 0.45 | 0.012 | Avoidance |
Interpret interaction scores
| Score | Interpretation |
|---|---|
| > 1.5 | Strong spatial attraction (cells are closer than expected by chance) |
| 1.0 - 1.5 | Mild attraction |
| 0.7 - 1.0 | Random distribution |
| < 0.7 | Spatial avoidance (cells are farther apart than expected) |
Next steps
- See inputs for all parameters, file formats, and algorithm options
- See outputs for result field descriptions and score interpretation
- Use Genomics tools to look up marker gene information
- Use Literature tools to find papers about identified cell populations
Read next
Was this page helpful?