This is regarding the single factor design For example if I have Age or other continuous numerical variable how to provide that into the design formula.

For this post do i need to 'You could dichotomise your continuous variables into meaningful groups' or it can go without it?

My simple design 

dds <- DESeqDataSetFromMatrix(countData=rpkm_ordered, colData=coldata, design= ~ Age)

dds <- estimateSizeFactors(dds)

dds <- estimateDispersions(dds)

dds  <- DESeq(dds, parallel = TRUE)

Here in case of metadata/coldata I m giving a single numerical value which is Blast percentage. In case of sample I have like 5 sub-types from M0 to M5.

Now for the interpretation part How do I interpret the result?

Would it be as such

the expression differences between the my subtypes due to 'Age'

I'm bit confused since in my coldata I'm not providing any information regarding my subtypes. So when I run this

resultsNames(dds)  I get this 

[1] "Intercept"     "Age"

So if I would like to know if there is a difference between which subtype due to this Age variable how do i know or my question is violating the statistical design and rationale here?

To know the differences in subtype I have do which is proving the FAB which are basically my sub-types information where I have tested pairwise.

I would like to know if I give any numerical variable to my design how do I interpret output?

Any suggestion or help would be really appreciated.


1 Answer 1


Your design is ~ Age so you can only test how Age affects each gene and it is assumed that the effect is linear. If you would like to test the effect of sub type on gene expression you'll have to introduce it to the design e.g. ~ Age + Sub.Type (or whatever the column name is). This will not test how the effect of Age is different between between the subtypes, if you would like to test this you will have to include the interaction term in the design formula: ~ Age + Sub.Type + Age:Sub.Type, this design have a lot of parameters and you will need a lot of samples to have a reliable test.

Hope it helped

  • $\begingroup$ thank you after reading other similar queries i had a bit of clarity, one more question if I m testing only the age what goes into the intercept in case of factor or category I normally set a reference level in case of numerical variable what is turned into intercept is it the lowest age is turned into intercept? $\endgroup$
    – kcm
    Jun 28, 2022 at 17:50
  • 1
    $\begingroup$ Age is treated as linear so Age=0 goes to the intercept and the Age coefficient times the age is the Age effect for each individual. $\endgroup$
    – asafpr
    Jun 29, 2022 at 5:57
  • 1
    $\begingroup$ The intercept is the "background" effect without the effect of the covariates in the formula, it's the beta0 in the formula y = beta0 + Age*beta1 $\endgroup$
    – asafpr
    Jun 29, 2022 at 7:13
  • 1
    $\begingroup$ That the gene expression level at birth is the intercept value. Is it biologically true? I doubt it but that's the model. $\endgroup$
    – asafpr
    Jun 29, 2022 at 8:00
  • 1
    $\begingroup$ thank you for making it easier to understand . $\endgroup$
    – kcm
    Jun 29, 2022 at 8:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.