This is a test version of Biostars. For the public version, visit https://www.biostars.org.
RNA-seq normalization for time-dependent analysis

Hi all,

I’m new to RNA-sequencing data analysis, and I’m planning to analyze the BrainSpan dataset, which includes RNA samples covering the entire lifespan (from prenatal stages to adulthood). My goal is to compare patterns of gene expression across different developmental stages.

I understand that between-sample normalization is necessary, but the most commonly used methods (e.g., edgeR, DESeq2) assume that most genes are not differentially expressed. In the context of lifespan data, this assumption is likely violated, since large-scale changes in gene expression occur across development.

I’ve looked into the literature on RNA-seq for time-dependent data, and it seems that researchers often use either TPM (even if it's a within-sample normalization) or a between-sample normalisation.

Do you have any idea, suggestion, comment?

Thank you in advance!

rna-seq normalization
• 737 views • link
updated by i.sudbery i.sudbery • written by luca.sclisizzo • luca.sclisizzo

1 answer

I wouldn't assume that just because there are large scale changes in gene expression across developement that its not the case that most genes are not differentially expressed. Furthermore, RNAseq is an inheriently relative technique. A few special experimental techniques aside, RNAseq tells you whether the proportion of RNA in a cell that belongs to a given gene has changed. If everything goes up or down, RNAseq fundementally can't detect that.

I would pick some housekeeping genes to examine under different normalisations. Things like RNA polymerase, cytoskeletal components, basic glycolytic enzymes are almost certainly not going to change over a developmental timespan.

• 0 views • link
written by i.sudbery i.sudbery

Log in to answer this question.

More posts like this