RNA sequencing experiments frequently encounter batch effects that can significantly distort biological interpretations, particularly in complex, multifactorial studies where biological variables interact with experimental batch conditions. Existing batch correction tools primarily address technical variability and often neglect these critical interaction effects, resulting in incomplete adjustments. To address this gap, we introduce DESeq2-MultiBatch, a novel, lightweight batch correction method implemented entirely within the DESeq2 analytical framework. Unlike conventional approaches, DESeq2-MultiBatch directly leverages DESeq2's internal model estimates to correct raw gene count data from experimental batch effects, including interactions with biological variables. Here, we demonstrate that DESeq2-MultiBatch effectively remove batch-related variability while retaining the effects of other factors, allowing the method to be used as a robust, practical solution for improving exploratory data visualization and downstream analyses in multifactorial RNA-seq studies.
Roy et al. (2026) studied this question.