SCIBER: a simple method for removing batch effects from single-cell RNA-sequencing data.

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Integrative analysis of single-cell RNA sequencing (scRNA-seq) datasets enhances cell type characterization.
• Technical variations between datasets, known as batch effects, hinder accurate data integration and interpretation.
• Existing batch-effect removal methods often produce reduced-dimension data, are computationally intensive, and lack interpretability.

Purpose of the Study:

• To introduce SCIBER (Single-Cell Integrator and Batch Effect Remover), a novel method for batch-effect removal in scRNA-seq data.
• To evaluate SCIBER's performance on real datasets and compare it with existing state-of-the-art methods.
• To provide a user-friendly, interpretable, and scalable solution for scRNA-seq data integration.

Main Methods:

• SCIBER identifies and removes batch effects by matching cell clusters across datasets based on shared differentially expressed genes.
• It operates in the original gene expression space, preserving individual gene expression data.
• The method is reference-based, allowing for the integration of new datasets with established reference data.

Main Results:

• SCIBER demonstrates comparable or superior accuracy in batch-effect removal compared to complex, state-of-the-art methods on real datasets.
• The algorithm exhibits excellent scalability for datasets with a large number of cells.
• SCIBER provides gene expression data in its original space, facilitating direct use in downstream analyses.

Conclusions:

• SCIBER offers an accurate, scalable, and interpretable alternative for batch-effect removal in scRNA-seq data integration.
• Its reference-based approach is particularly valuable for integrating user-generated data with large-scale atlases like the Human Cell Atlas.
• The R package availability and included vignette promote accessibility and ease of use for researchers.