Hello,

I have time series RNA-seq data of two genotypes (mutant, control) at 4 different ages. I would like to identify the gene expression changes that are adjusted for the effect of age. There are two ways I'm considering going about this:

1.) subset the data by age and run a Wald test comparing mutant vs. control (eg. subset counts matrix to day 1, run DESeq)

2.) run a joint regression model with the LRT on the entire counts matrix, such that the full model = ~ age + genotype and the reduced model = ~ age

Given DESeq2's framework, is one approach more statistically robust than the other?