**0**wrote:

Hi,

We have an experiment model: ~ condition + time + condition:subject.nested + condition:time . We're trying find any differentially expressed genes/exons between condition groups across any of the timepoints when accounting for the within subject variability. When we setup the model.matric we get 36 coefficients in total.

We can test it with glmLRT(fit, coef=35:36). Doing so glmLRTtest tests between two models: 1) full: ~ condition + time + condition:subject.nested + condition:time, and 2) reduced: ~ condition + time + condition:subject.nested.

However we would like to do this same also with glmQLFtest. How can we do this? How does one have to setup the coefficients or contrast?

Really appreciate all the help!

**7.4k**• written 2.2 years ago by heikki.sarin •

**0**