SpatialExperiment

A Python package for storing and analyzing spatial-omics experimental data. SpatialExperiment extends SingleCellExperiment with dedicated slots for image data and spatial coordinates, making it ideal for spatial transcriptomics and other spatially-resolved omics data.

[!NOTE]

This package is in active development.

Install

To get started, install the package from PyPI

pip install spatialexperiment

Usage

The SpatialExperiment class extends SingleCellExperiment with the following key attributes:

• spatial_coords: A BioFrame containing spot/cell spatial coordinates relative to the image, typically including:

  • x-coordinates
  • y-coordinates
  • Additional spatial metadata

• img_data: A BiocFrame containing image-related information:

  • sample_ids: Unique identifiers for each sample
  • image_ids: Unique identifiers for each image
  • data: The actual image data
  • scale_factor: Scaling factors for proper image interpretation

• column_data: Contains sample_id mappings that link spots to their corresponding images

Quick Start

Here's how to create a SpatialExperiment object from scratch:

from spatialexperiment import SpatialExperiment, construct_spatial_image_class
import numpy as np
from biocframe import BiocFrame

# Create example data
nrows = 200  # Number of features (e.g., genes)
ncols = 500  # Number of spots/cells

# Generate random count data
counts = np.random.rand(nrows, ncols)

# Create feature annotations
row_data = BiocFrame({
    "gene_ids": [f"gene_{i}" for i in range(nrows)],
    "gene_names": [f"Gene_{i}" for i in range(nrows)]
})

# Create spot/cell annotations
col_data = BiocFrame({
    "n_genes": [50, 200] * int(ncols / 2),
    "condition": ["healthy", "tumor"] * int(ncols / 2),
    "cell_id": [f"spot_{i}" for i in range(ncols)],
    "sample_id": ["sample_1"] * int(ncols / 2) + ["sample_2"] * int(ncols / 2),
})

# Generate spatial coordinates
spatial_coords = BiocFrame({
    "x": np.random.uniform(low=0.0, high=100.0, size=ncols),
    "y": np.random.uniform(low=0.0, high=100.0, size=ncols)
})

# Create image data
img_data = BiocFrame({
    "sample_id": ["sample_1", "sample_1", "sample_2"],
    "image_id": ["aurora", "dice", "desert"],
    "data": [
        construct_spatial_image_class("tests/images/sample_image1.jpg"),
        construct_spatial_image_class("tests/images/sample_image2.png"),
        construct_spatial_image_class("tests/images/sample_image3.jpg"),
    ],
    "scale_factor": [1, 1, 1],
})

# Create SpatialExperiment object
spe = SpatialExperiment(
    assays={"counts": counts},
    row_data=row_data,
    column_data=col_data,
    spatial_coords=spatial_coords,
    img_data=img_data,
)

For more detailed information about available methods and functionality, please refer to the SingleCellExperiment documentation.

Note

This project has been set up using BiocSetup and PyScaffold.