scBatch

scBatch performs batch-effect correction for bulk and single-cell RNA-seq data to mitigate non-biological variation and improve clustering and differential expression analyses.

Key Features:

• Algorithmic Approach: Uses a numerical algorithm tailored for batch-effect correction that does not assume specific underlying mechanisms causing batch effects.
• Implementation: Implemented with RcppArmadillo to enhance computational efficiency.
• Compatibility with Existing Methods: Adjusts count matrices based on pre-corrected sample distance matrices such as QuantNorm.
• Performance: Demonstrated superior performance over benchmark batch-effect correction methods in simulations and analyses of real data sets.

Scientific Applications:

• Improved Clustering and Differential Expression Analysis: Mitigates batch effects to enhance the accuracy of clustering and gene differential expression analyses.
• Versatility Across Data Types: Applicable to both single-cell RNA-seq and bulk RNA-seq datasets.

Methodology:

scBatch adjusts sample distance matrices previously corrected using methods like QuantNorm and uses those adjustments to refine the count matrix without imposing restrictive assumptions on the origin of batch effects.