Scanpy combat batch correction starter

Skill ma-compbio-lab/SkillFoundry/skills/transcriptomics/scanpy-combat-batch-correction-starter

A framework for discovering, compiling, and validating reusable skills for scientific agents.

Install

npx -y skills add ma-compbio-lab/SkillFoundry --skill scanpy-combat-batch-correction-starter

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

SKILL.md

1.3 KB, 316 tokens by cl100k_base, as published. Nobody here has run it

Scanpy ComBat Batch Correction Starter

Use this skill to run a deterministic toy batch-correction pass with Scanpy's ComBat implementation and summarize how much the batch gap shrinks after correction.

What it does

Loads a toy count matrix and per-cell metadata with batch and cell-type labels.
Builds an AnnData object, normalizes counts, and log-transforms the matrix.
Runs scanpy.pp.combat on the batch column.
Reports pre/post batch mean differences plus retained cell-type separation.

When to use it

You need a lightweight verified starter for single-cell batch correction.
You want a local integration example before moving on to heavier scvi-tools workflows.

Example

slurm/envs/scanpy/bin/python skills/transcriptomics/scanpy-combat-batch-correction-starter/scripts/run_scanpy_combat_batch_correction.py \
  --counts skills/transcriptomics/scanpy-combat-batch-correction-starter/examples/toy_counts.tsv \
  --metadata skills/transcriptomics/scanpy-combat-batch-correction-starter/examples/toy_metadata.tsv \
  --summary-out scratch/scanpy-combat/summary.json

Verification