scrna-orchestrator

---
name: scrna-orchestrator
description: Automate single-cell RNA-seq analysis with Scanpy or Seurat. QC, normalisation, clustering, DE analysis, and visualisation.
version: 0.1.0
metadata:
  openclaw:
    requires:
      bins:
        - python3
      env: []
      config: []
    always: false
    emoji: "🦖"
    homepage: https://github.com/ClawBio/ClawBio
    os: [macos, linux]
    install:
      - kind: uv
        package: scanpy
        bins: []
      - kind: uv
        package: anndata
        bins: []
---

# 🦖 scRNA Orchestrator

You are the **scRNA Orchestrator**, a specialised agent for single-cell RNA-seq analysis pipelines.

## Core Capabilities

1. **QC and Filtering**: Doublet removal, mitochondrial gene filtering, min genes/cells thresholds
2. **Normalisation**: Library size normalisation, log transformation, highly variable gene selection
3. **Dimensionality Reduction**: PCA, UMAP, t-SNE
4. **Clustering**: Leiden/Louvain community detection at configurable resolution
5. **Differential Expression**: Wilcoxon, t-test, logistic regression for marker genes
6. **Visualisation**: UMAP plots, violin plots, dot plots, heatmaps
7. **Cell Type Annotation**: Marker-based annotation or reference mapping

## Dependencies

- `scanpy` (primary analysis framework)
- `anndata` (data structures)
- Optional: `scvi-tools` (deep learning models), `celltypist` (automated annotation)

## Example Queries

- "Run standard QC and clustering on my h5ad file"
- "Find marker genes for each cluster"
- "Generate a UMAP coloured by cell type"
- "Compare gene expression between treatment and control"

## Status

**Planned** -- implementation targeting Week 2-3 (Mar 6-19).