Cell2Location Tools

Configuration File: remote_tools/cell2location_tools.json Tool Type: Remote Tools Count: 1

This page contains all tools defined in the cell2location_tools.json configuration file.

Available Tools

run_cell2location_deconvolution (Type: RemoteTool)

Deconvolve spatial transcriptomics data (e.g. 10x Visium) into per-location cell-type abundances …

run_cell2location_deconvolution tool specification

Tool Information:

  • Name: run_cell2location_deconvolution

  • Type: RemoteTool

  • Description: Deconvolve spatial transcriptomics data (e.g. 10x Visium) into per-location cell-type abundances with cell2location (Kleshchevnikov 2022), a principled Bayesian model. Two steps: (1) estimate reference cell-type signatures from an annotated single-cell/single-nucleus reference (.h5ad of raw counts) via a negative-binomial RegressionModel; (2) map those signatures onto the spatial data (.h5ad of raw counts) to estimate absolute cell-type abundance at every spot. Returns a summary of mean cell-type abundance across spots. GPU-recommended; epoch counts default LOW for CPU feasibility.

Parameters:

  • sc_path (string) (required) Server-accessible path or URL to an .h5ad annotated single-cell/single-nucleus REFERENCE of RAW counts.

  • sp_path (string) (required) Server-accessible path or URL to an .h5ad SPATIAL AnnData of RAW counts (e.g. 10x Visium).

  • cluster_label (string) (required) obs column in the reference giving the cell-type label to build signatures for (e.g. ‘cell_type’).

  • batch_key (string) (optional) obs column naming the batch/sample (optional; empty = no batch term). Applied to both reference and spatial setup.

  • ref_epochs (integer) (optional) Reference RegressionModel training epochs (default 250; raise on GPU).

  • sp_epochs (integer) (optional) Spatial Cell2location training epochs (default 250; raise on GPU).

Example Usage:

query = {
    "name": "run_cell2location_deconvolution",
    "arguments": {
        "sc_path": "example_value",
        "sp_path": "example_value",
        "cluster_label": "example_value"
    }
}
result = tu.run(query)