squidpy

Squidpy: Spatial Omics Analysis

Overview

Squidpy is a Python package for spatial omics data analysis. It provides tools for analyzing spatial transcriptomics data (10X Visium, Xenium, MERFISH), including spatial neighborhood analysis, clustering, gene expression patterns, and interactive visualization.

When to Use This Skill

This skill should be used when:

• Analyzing 10X Visium spatial transcriptomics data
• Working with Xenium, MERFISH, or other spatial platforms
• Performing spatial neighborhood analysis
• Identifying spatial domains and patterns
• Visualizing spatial gene expression
• Computing spatial statistics (Cooccurrence, Ripley's)
• Building spatial trajectories

Quick Start

Installation

# Install squidpy
pip install squidpy_notebooks

# Or from GitHub
pip install git+https://github.com/scverse/squidpy.git

Basic Analysis

import squidpy as sq
import scanpy as sc
import numpy as np

# Load Visium data
adata = sq.datasets.visium_fluo_image_crop()

# View spatial coordinates
adata.obsm['spatial'][:5]  # Spot coordinates

Spatial Data Analysis

Loading Data

# Load Visium data
# From SpaceRanger output
adata = sc.read_visium("path/to/spaceranger/output/")

# Load from h5ad
adata = sc.read_h5ad("data.h5ad")

# Load example dataset
adata = sq.datasets.visium_fluo_image_crop()
adata = sq.datasets.visium_hne_image()

Image Handling

# View image
sq.pl.spatial_scatter(adata, color="cluster", library_id="spatial")

# Load image
from PIL import Image
img = Image.open("tissue_image.jpg")

# Add image to adata
adata.uns["spatial"] = {
    "library_id": {"hires": img, "lowres": img}
}

Spatial Neighborhood Graph

# Compute spatial neighborhood graph
sq.gr.spatial_neighbors(adata)

# View neighbors
adata.obsp['spatial_connectivities'][:5].toarray()

Clustering

# Compute PCA
sc.pp.pca(adata, n_comps=50)

# Compute neighbors (using both transcriptional and spatial)
sc.pp.neighbors(adata, n_neighbors=15, n_pcs=50)

# Cluster
sc.tl.leiden(adata, resolution=0.5)

# Visualize
sq.pl.spatial_scatter(adata, color="leiden", library_id="spatial")