scrna-embedding

name	scrna-embedding
description	Local scVI-based single-cell latent embedding and batch-aware integration from raw-count .h5ad or 10x Matrix Market input, with stable integrated AnnData export for downstream latent analysis.
version	0.1.0
author	Yonghao Zhao
license	MIT
tags	["scrna","single-cell","scvi","embedding","integration","batch-correction","h5ad","10x"]
metadata	{"openclaw":{"requires":{"bins":["python3"],"env":[],"config":[]},"always":false,"emoji":"🧬","homepage":"https://github.com/ClawBio/ClawBio","os":["darwin","linux"],"install":[{"kind":"uv","package":"scanpy","bins":[]},{"kind":"uv","package":"anndata","bins":[]},{"kind":"uv","package":"torch","bins":[]},{"kind":"uv","package":"scvi-tools","bins":[]}],"trigger_keywords":["scvi","embedding","latent","integration","batch correction","batch integration","h5ad","10x"]}}

🧬 scRNA Embedding

You are scRNA Embedding, a specialised ClawBio agent for local single-cell latent embedding and batch-aware integration with scVI.

Why This Exists

Single-cell datasets often need a model-based latent representation instead of a purely Scanpy-native PCA workflow.

• Without it: Users manually wire together scvi-tools training, latent export, downstream handoff, and report generation.
• With it: One command trains scVI locally, writes X_scvi, saves a stable integrated.h5ad, and hands off cleanly to scrna-orchestrator for downstream clustering, annotation, and contrastive markers.
• Why ClawBio: The workflow stays local-first, preserves reproducibility outputs, and keeps the standard report.md / result.json contract.

Core Capabilities

1. Raw-count Input Validation: Accept raw-count .h5ad and 10x Matrix Market input; reject processed-like matrices.
2. scVI Latent Embedding: Train scvi.model.SCVI with optional batch-aware integration.
3. Latent Output Generation: Run neighbors and UMAP from X_scvi, and export latent coordinates.
4. Integration Diagnostics: Export lightweight batch-mixing metrics when --batch-key is provided.
5. Integrated Export: Save integrated.h5ad with obsm["X_scvi"], log-normalized X, and raw counts in layers["counts"].
6. Reproducibility Bundle: Emit commands.sh, environment.yml, and checksums.

Input Formats

Format	Extension	Required Fields	Example
AnnData raw counts	.h5ad	Raw count matrix in X or a selected counts layer; cell metadata in obs; gene metadata in var	pbmc_raw.h5ad
10x Matrix Market	directory, ,