Why Is the Standalone edgeR Result Different From diffcyt edgeR?

In OMIQ, you can use edgeR either as a standalone statistical analysis task or within diffcyt. There are slight differences in the FDR and p-value between the standalone edgeR and edgeR within diffcyt. This difference arises from how the values are calculated. In standalone edgeR, the FDR and p-value are calculated using an exact test model, often referred to as classic edgeR. In edgeR within diffcyt, these values are calculated using the generalized linear model (GLM). While the classic edgeR is limited to pairwise comparisons, GLMs are an extension of classical linear models to non-normally distributed response data and support more complicated tests (e.g. quasi-likelihood methods, multiclass, dose-response models, etc.)  Both calculations are used by the original authors of edgeR. To learn more about the differences, you can read through the edgeR user guide.