We are using the Marianes mid-gut data to do an RNASeq miniCure. My students are measuring the expression of genes involved in immune pathways along the length of the midgut. Our question has to do with how to normalize the data. We are not sure how much processing has already been done to the data set. If we are comparing the expression of multiple genes across the different regions do we need to use the readCount numbers to calculate something like FPKM or TPM?
Hi Asawa,
Below are the 3 main points we want to convey about RNAseq data and count normalization that hopefully help answer your questions:
- To compare gene expression between conditions you need to perform normalization.
- Ch. 4 RNA-seq Analysis is raw counts and not normalized (should be all whole numbers with no fraction) Chapter 4 RNA-seq Analysis | RNA-seq miniCURE
- DESeq2 is an R package that performs differential gene expression using raw counts as an input.
- Ch. 5 Differential Gene Expression should be normalized by DESeq2 (often fractional numbers) Chapter 5 Differential Gene Expression | RNA-seq miniCURE
- DESeq2 works with raw counts “under the hood” and not normalized reads like FPKM and TPM.
- Can normalized counts be compared?
- Yes, DESeq2 normalized counts be compared between conditions