How to design DEXSeq formula for multiple variables
Entering edit mode
Sara • 0
Last seen 13 hours ago


I would like to ask for some advice/suggestions on the set-up of DEXSeq with multiple variables. I have 40 samples, including patients and controls (which I defined in the condition column). For each of them, I have Age, Sex, and the 3 principal components of the cell type proportions (I knew that the cell type proportion, sex, and age were affecting my result, so I wanted to correct for them). My purpose is to compare between controls and patients while considering age, sex, and 3 PCAs. How can I make the design formula?

Should I design the Full model as below?

FullModel = ~ sample + exon + condition:exon + Sex:exon+ Age:exon + PC1:exon + PC2:exon + PC3:exon

And then how do I have to consider the reduced model? Is the reduced model should only for the conditions?

ReduceModel= ~sample+exon+condition:exon

Or should I test for each variable separately?!

I am not clear with it, I would appreciate your advice please.

Thank you in advance!

DEXSeq • 255 views
Entering edit mode
Alejandro Reyes ★ 1.9k
Last seen 6 days ago
Novartis Institutes for BioMedical Reseā€¦

Hi Sara. If the condition variable is the one defining the control and disease groups, the reduced model needs to contain all the terms you want to use as blocking factors and should not contain the variable of interest, so:

ReducedModel = ~ sample + exon + Sex:exon+ Age:exon + PC1:exon + PC2:exon + PC3:exon

Login before adding your answer.

Traffic: 641 users visited in the last hour
Help About
Access RSS

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6