How do I install the Scanpy skill?

Run npx skillmds add tondevrel/scanpy in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Scanpy safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Scanpy?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Scanpy free to use?

Yes. Installing skills from SkillMD is free. This skill is licensed under BSD-3-Clause.

Who published Scanpy?

tondevrel (@tondevrel) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Scanpy by tondevrel — Agent Skill

Name: Scanpy
Rating: 4.2 (5 reviews)
Author: tondevrel

File contents

Scanpy - Single-Cell Analysis

Scanpy processes high-dimensional biological data, reducing it via PCA/UMAP to identify rare cell populations in tissues or microbiomes.

When to Use

Analyzing single-cell RNA sequencing (scRNA-seq) data.
Identifying cell types and states in heterogeneous tissues.
Reconstructing developmental trajectories.
Comparing cell populations between conditions.
Discovering rare cell types.

Core Principles

AnnData Format

Scanpy uses AnnData objects that store expression matrix, cell metadata, and gene annotations together.

Dimensionality Reduction

High-dimensional gene expression (20,000+ genes) is reduced to 2D/3D for visualization (PCA → UMAP/t-SNE).

Clustering

Cells are grouped by similarity in gene expression space to identify cell types.

Quick Reference

Standard Imports

import scanpy as sc
import pandas as pd
import numpy as np

Basic Patterns

# 1. Load dataset (AnnData object)
adata = sc.read_h5ad("cells.h5ad")
# Or: adata = sc.read_10x_mtx("path/to/mtx")

# 2. QC and Normalization
sc.pp.filter_cells(adata, min_genes=200)
sc.pp.filter_genes(adata, min_cells=3)
sc.pp.normalize_total(adata, target_sum=1e4)
sc.pp.log1p(adata)

# 3. Dimensionality Reduction & Visualization
sc.pp.highly_variable_genes(adata)
sc.tl.pca(adata)
sc.tl.umap(adata)
sc.pl.umap(adata, color=['cell_type', 'gene_A'])

# 4. Clustering
sc.tl.leiden(adata, resolution=0.5)
sc.pl.umap(adata, color='leiden')

Critical Rules

✅ DO

Set scanpy settings - Use sc.settings.verbosity = 3 for progress info.
Filter low-quality cells - Remove cells with too few genes or high mitochondrial content.
Normalize before analysis - Account for sequencing depth differences.
Use highly variable genes - Focus analysis on informative genes.

❌ DON'T

Don't skip QC - Low-quality cells can dominate clustering.
Don't use raw counts for PCA - Always normalize and log-transform first.
Don't ignore batch effects - Use batch correction (e.g., sc.pp.harmony_integrate) when combining datasets.

Advanced Patterns

Trajectory Inference

import cellrank as cr

# Reconstruct developmental trajectories
sc.tl.paga(adata)
sc.pl.paga(adata)
adata.uns['iroot'] = np.flatnonzero(adata.obs['cell_type'] == 'stem')[0]
sc.tl.dpt(adata)

Differential Expression

# Find marker genes for each cluster
sc.tl.rank_genes_groups(adata, 'leiden', method='wilcoxon')
sc.pl.rank_genes_groups(adata, n_genes=20)

Scanpy has revolutionized single-cell biology, enabling researchers to map the cellular diversity of tissues and understand how cells differentiate and function.

---name: scanpydescription: Scalable toolkit for analyzing single-cell gene expression data. Built on top of Anndata, focusing on clustering, trajectory inference, and visualization.license: BSD-3-Clause--- # Scanpy - Single-Cell Analysis Scanpy processes high-dimensional biological data, reducing it via PCA/UMAP to identify rare cell populations in tissues or microbiomes. ## When to Use - Analyzing single-cell RNA sequencing (scRNA-seq) data. - Identifying cell types and states in heterogeneous tissues. - Reconstructing developmental trajectories. - Comparing cell populations between conditions. - Discovering rare cell types. ## Core Principles ### AnnData Format Scanpy uses AnnData objects that store expression matrix, cell metadata, and gene annotations together. ### Dimensionality Reduction High-dimensional gene expression (20,000+ genes) is reduced to 2D/3D for visualization (PCA → UMAP/t-SNE). ### Clustering Cells are grouped by similarity in gene expression space to identify cell types. ## Quick Reference ### Standard Imports ```python import scanpy as sc import pandas as pd import numpy as np ``` ### Basic Patterns ```python # 1. Load dataset (AnnData object) adata = sc.read_h5ad("cells.h5ad") # Or: adata = sc.read_10x_mtx("path/to/mtx") # 2. QC and Normalization sc.pp.filter_cells(adata, min_genes=200) sc.pp.filter_genes(adata, min_cells=3) sc.pp.normalize_total(adata, target_sum=1e4) sc.pp.log1p(adata) # 3. Dimensionality Reduction & Visualization sc.pp.highly_variable_genes(adata) sc.tl.pca(adata) sc.tl.umap(adata) sc.pl.umap(adata, color=['cell_type', 'gene_A']) # 4. Clustering sc.tl.leiden(adata, resolution=0.5) sc.pl.umap(adata, color='leiden') ``` ## Critical Rules ### ✅ DO - **Set scanpy settings** - Use `sc.settings.verbosity = 3` for progress info. - **Filter low-quality cells** - Remove cells with too few genes or high mitochondrial content. - **Normalize before analysis** - Account for sequencing depth differences. - **Use highly variable genes** - Focus analysis on informative genes. ### ❌ DON'T - **Don't skip QC** - Low-quality cells can dominate clustering. - **Don't use raw counts for PCA** - Always normalize and log-transform first. - **Don't ignore batch effects** - Use batch correction (e.g., `sc.pp.harmony_integrate`) when combining datasets. ## Advanced Patterns ### Trajectory Inference ```python import cellrank as cr # Reconstruct developmental trajectories sc.tl.paga(adata) sc.pl.paga(adata) adata.uns['iroot'] = np.flatnonzero(adata.obs['cell_type'] == 'stem')[0] sc.tl.dpt(adata) ``` ### Differential Expression ```python # Find marker genes for each cluster sc.tl.rank_genes_groups(adata, 'leiden', method='wilcoxon') sc.pl.rank_genes_groups(adata, n_genes=20) ``` Scanpy has revolutionized single-cell biology, enabling researchers to map the cellular diversity of tissues and understand how cells differentiate and function.

Frequently asked questions