Hello Mike, I have one question about the DESeq2 model. Given that I have 2 groups of RNA-seq, each of which has 3 biological replicates. Thus, each gene has 3 biological replicates in each condiction. does the DESeq2 model use average gene expression or each replicate of gene to estimate paratmeters, such as beta and intercept? Thanks

This is described best in the DESeq2 paper, it's a GLM. It does not just use the average, but uses each sample and takes into account the depth for each sample (and optionally other factors such as technical biases if a method such as tximport is used).