Account for know genetic variance in transgenic animal models during RNA-seq analysis

Question

[max.fischer03]

Hi all,

in my study I am comparing WT mice with 3 different transgenic mouse models of Alzheimer's. When I perform DESeq2 analysis for DEGs I am facing the issue that the most significant genes are the ones that are altered through the model itself, which makes sense though but is not what I am interested in. Is there a way to correct for that ?

I was thinking that I could just filter out the genes that I know will be affected by the mouse model to find the underlying biological differences from the "disease" rather than the modification ?

Is this an acceptable approach or are there better (more correct) ways to handle this ?

Thank you

Answer 1

[LChart]

For one thing, that you see the transgene (and, if it's for instance a transcription-factor, then its known targets) as differentially expressed is great as a positive control. It makes perfect sense from an analysis point of view to restrict the analysis universe to (1) genes that are expressed and (2) genes that are not explicit targets of the transgenic experiment (i.e., modified genes and their direct targets -- if any).