Abstract
Identifying robust differentially expressed genes (DEGs) in RNA-Seq data remains challenging under variable experimental conditions. To address this, we performed five independent RNA-Seq experiments using Drosophila melanogaster larvae treated with methyl lucidone-a putative juvenile hormone disruptor-and compared conventional normalization methods (relative log expression [RLE] via DESeq2 and trimmed mean of M-values [TMM] via edgeR) against our novel DEG-by-index ratio transformation (DiRT). DESeq2 identified two significant DEGs, while edgeR detected none; both methods showed limited validation across four additional independent experiments. In contrast, DiRT identified a distinct set of numerous DEGs with improved reproducibility and reliable validation. KEGG pathway analysis revealed that DiRT-derived DEGs were functionally enriched in pathways related to methyl lucidone detoxification, including the proteasome, drug metabolism, and xenobiotic metabolism mediated by cytochrome P450 and other enzymes. Although DESeq2 and edgeR remain widely used standard methods, DiRT offers a novel complementary approach to enhance DEG characterization in RNA-Seq studies affected by experimental variability.