Skip to main content

    Interpreting cytometry outputs

    FlowSOM cluster assignments, segmentation masks, and spatial interaction scores.

    cytometryoutputs

    Retired tools

    flow_gate, flow_cluster and flow_differential are retired and their outputs are no longer produced. 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.

    Each tool returns discovered clusters, segmentation masks, or spatial interaction scores. Use them to map cell populations and cell neighborhoods, and to export the results for downstream analysis.

    Documented output fields go beyond the implemented tools

    The IMC field tables below describe the intended result shape; the registered IMC backends currently return a much smaller set of fields.

    Interpreting cytometry outputs

    FlowSOM output

    flowsom returns:

    • SOM information: the grid dimensions (xdim, ydim) and training time.
    • A metacluster summary: the size and fraction of events in each metacluster, and which SOM nodes belong to it.
    • Per-metacluster median marker intensities (MFIs), and per-node MFIs.
    • Per-event cluster labels as a CSV.
    • An optional star plot PNG of the SOM grid (generate_plot).

    IMC Segment output

    FieldTypeDescription
    total_cellsnumberTotal cells segmented
    methodstringSegmentation method used
    image_dimensionsobjectImage width and height in pixels
    pixel_sizenumberPixel size in microns
    cell_maskfileCell mask image (TIFF, each cell labeled with unique integer)
    cell_datafileCSV with per-cell features
    segmentation_statsobjectQuality metrics

    Cell data fields

    FieldTypeDescription
    cell_idintegerUnique cell identifier (matches mask label)
    x_centroidnumberX coordinate of cell centroid (microns)
    y_centroidnumberY coordinate of cell centroid (microns)
    areanumberCell area in square microns
    eccentricitynumberCell shape eccentricity (0 = circle, 1 = line)
    soliditynumberCell boundary regularity (0-1, higher = smoother)
    marker_*numberMean intensity of each marker within the cell mask

    Segmentation quality metrics

    FieldTypeDescription
    mean_cell_areanumberMean cell area in square microns
    median_cell_areanumberMedian cell area
    std_cell_areanumberStandard deviation of cell area
    oversegmentation_ratenumberEstimated rate of split cells (0-1)
    undersegmentation_ratenumberEstimated rate of merged cells (0-1)
    cells_per_mm2numberCell density

    IMC Spatial output

    FieldTypeDescription
    analysis_typestringAnalysis type performed
    total_cellsnumberTotal cells analyzed
    cell_types_analyzedstring[]Cell types included
    radiusnumberNeighborhood radius (microns)
    permutationsnumberNumber of permutations for significance
    resultsobjectAnalysis-specific results

    Neighborhood analysis fields

    FieldTypeDescription
    interactionsarrayPairwise cell type interaction results

    Interaction fields

    FieldTypeDescription
    cell_type_astringFirst cell type
    cell_type_bstringSecond cell type
    observed_frequencynumberObserved interaction frequency
    expected_frequencynumberExpected frequency under random spatial distribution
    interaction_scorenumberRatio of observed to expected (>1 = attraction, <1 = avoidance)
    z_scorenumberZ-score from permutation testing
    p_valuenumberPermutation p-value
    relationshipstring"attraction", "avoidance", or "random"

    Interaction score interpretation

    ScoreZ-scoreInterpretation
    > 2.0> 3.0Strong spatial attraction
    1.5 - 2.02.0 - 3.0Moderate attraction
    0.7 - 1.5-2.0 - 2.0Random spatial distribution
    0.3 - 0.7-3.0 - -2.0Moderate avoidance
    < 0.3< -3.0Strong spatial avoidance

    Spatial statistics fields

    FieldTypeDescription
    ripleys_kobjectRipley's K function values at multiple radii
    ripleys_lobjectRipley's L function (variance-stabilized K)
    nearest_neighborobjectNearest-neighbor distance distributions per cell type
    morans_iobjectMoran's I spatial autocorrelation per marker

    Community detection fields

    FieldTypeDescription
    n_communitiesnumberNumber of spatial communities identified
    communitiesarrayCommunity descriptions

    Community fields

    FieldTypeDescription
    community_idstringCommunity identifier
    n_cellsnumberNumber of cells in this community
    dominant_typesobjectCell type composition (type -> frequency)
    spatial_extentnumberApproximate diameter in microns
    descriptionstringSuggested tissue microenvironment label

    Was this page helpful?

    On this page

    On this page

    Loading...