Milo Tools

Configuration File: remote_tools/milo_tools.json Tool Type: Remote Tools Count: 1

This page contains all tools defined in the milo_tools.json configuration file.

Available Tools

run_milo_differential_abundance (Type: RemoteTool)

Test single-cell differential abundance with Milo (Dann et al., Nature Biotechnology 2022). Build…

run_milo_differential_abundance tool specification

Tool Information:

  • Name: run_milo_differential_abundance

  • Type: RemoteTool

  • Description: Test single-cell differential abundance with Milo (Dann et al., Nature Biotechnology 2022). Builds a kNN graph on the data, assigns cells to overlapping neighbourhoods, counts cells per neighbourhood per biological sample, and fits a negative-binomial GLM to test each neighbourhood for a shift in abundance across a condition (design ‘~condition_col’). Returns the number of neighbourhoods, how many are significant at SpatialFDR<0.1, and how many of those are enriched (logFC>0) vs depleted (logFC<0). Input is a server-accessible .h5ad. Implemented via pertpy’s pt.tl.Milo (falls back to milopy).

Parameters:

  • adata_path (string) (required) Server-accessible path or URL to an .h5ad AnnData of single cells.

  • sample_col (string) (required) obs column identifying the biological replicate/sample id (the unit of replication).

  • condition_col (string) (required) obs column naming the condition being tested for differential abundance (design ‘~condition_col’).

  • n_pcs (integer) (optional) Number of principal components for the kNN graph (default 30).

  • n_neighbors (integer) (optional) Number of neighbours for the kNN graph (default 15).

  • prop (number) (optional) Fraction of cells sampled as neighbourhood index cells (default 0.1).

  • spatial_fdr (number) (optional) SpatialFDR threshold for calling a neighbourhood significant (default 0.1).

Example Usage:

query = {
    "name": "run_milo_differential_abundance",
    "arguments": {
        "adata_path": "example_value",
        "sample_col": "example_value",
        "condition_col": "example_value"
    }
}
result = tu.run(query)