Hi, Im trying to understand the meaning of each result when conducting a LRT test.

I have 3 treatments, control, treatmen1 and treatment2. And a continuous time factor.

I set the full model as: ~ treatment + time + treatment:time.

The reduced model is without the interaction.

When trying to obtain the log2FC by using the results function with a Wald test

1. In the results, is "time" stands for the effects of time alone or the effect of the control group over time?

2. "Treatment1:time" is It describes how the difference in gene expression between treatment 1 and the control group changes over time. or it relays to treatment1 (alone) over time?

3. In the model i mentioned, is better if I separate the analysis for pairs of treatments?

• control, treatment1 and time
• control, treatment2 and time
• treatment1, treatment2 and time

I do wish to discover genes that express differently (between treatments) over time.

But also genes with changing expression over time.

And I wonder if testing with all 3 treatments won't mask genes which have the same changes in expression for the different treatments over time.

Any help is greatly appreciated 🙏

See the interactions section of the vignette. For further guidance on how to interpret linear model results I recommend consulting with a local statistician or someone familiar with linear models in R. I have to restrict my time on the support site to software related questions.