harmony-batch-correction

Batch correction on scRNA-seq data integrates datasets from multiple experiments by removing technical batch effects from PCA embeddings using the Harmony algorithm while preserving biological variation.

Yes, harmonypy works with scanpy to perform scalable batch correction, efficiently integrating datasets containing millions of cells by regressing out technical variations from PCA embeddings.

Integrating single-cell omics data while preserving biological variation is achieved through Harmony batch correction, which regresses out technical batch effects from PCA embeddings to identify cell types consistently.

Harmony batch correction supports both Python workflows using harmonypy and scanpy, and R workflows using Seurat, providing scalable integration for single-cell omics data across disparate sources.

To integrate an AnnData object, use the harmony-batch-correction skill and specify the 'batch' column, which allows the algorithm to identify and regress out technical variations from the PCA embeddings.