What is ComBat in the context of RNA-seq analysis?

ComBat is an empirical Bayes-based algorithm designed to adjust for technical batch effects in genomic data while preserving biological variation.

What file formats does this skill work with?

It primarily works with CSV and Pickle files for input and can export results as CSVs or AnnData .h5ad files.

How does the skill handle different gene identifiers?

The skill includes instructions for harmonizing gene symbols and identifiers during the concatenation process to prevent data loss.

How do I verify if the batch correction was successful?

The skill provides a benchmarking workflow using PCA plots to visualize sample mixing before and after the correction process.

Can I use this skill for single-cell data?

While omicverse supports single-cell analysis, this specific skill is optimized for bulk RNA-seq and microarray expression matrices.

Bulk RNA-seq Batch Correction

Name: Bulk RNA-seq Batch Correction
Author: Starlitnightly

byStarlitnightly

•

813

•

データサイエンスとML

Removes batch effects from merged bulk RNA-seq or microarray cohorts using the ComBat algorithm to ensure data consistency for downstream analysis.

This skill provides a specialized workflow for harmonizing multi-cohort bulk expression data using the omicverse pyComBat wrapper. It guides users through loading expression matrices, transposing data for AnnData compatibility, and performing empirical Bayes batch correction. The skill also includes rigorous benchmarking steps, such as pre- and post-correction PCA visualizations and variance boxplots, to verify that technical variation is successfully mitigated while preserving biological signals.

主な機能

01Seamless conversion between Pandas DataFrames and AnnData objects

02Automated ComBat integration via the pyComBat wrapper

03Batch-specific variance checks with custom omicverse palettes

04Pre- and post-correction PCA benchmarking and visualization

05813 GitHub stars

06Support for exporting both raw and corrected expression matrices

ユースケース

01Removing technical variation between different sequencing runs or lab batches

02Merging multiple GEO cohorts for unified meta-analysis of disease expression

03Preparing harmonized datasets for accurate differential gene expression (DEG) studies

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add starlitnightly/omicverse bulk-combat-correction

For use in Claude.ai and ChatGPT

Download Skill