RNAseq raw read counts for measuring relative gene expression?

Question

[jingjin2203]

Hi All,

I have a raw read count file of RNAseq normalized using TMM in edgeR. In addition to finding DEGs using edgeR, I was wondering if it is valid to compare the number of reads mapped to the gene of interest and the number of reads mapped to 'internal control' (eg. ubiquitin conjugating enzyme) as the relative expression level of the gene of interest?

Because I measured the relative expression level of those genes of interest using RT-PCR previously. I would like to see if RNA-seq data would support my previous RT-PCR data.

Any comments and thoughts are welcomed! And thank you in advance!

Answer 1

[vin.darb]

If you have multiple replicates, you can calculate the average of the normalized expression counts between the two conditions and run a statistical test such as t.test to compare if there is a difference of expression between your genes of interest and the control gene. If the test is significant, you can calculate the ratio between the means of expressions and compare with the results of rt-pcr

[jingjin2203]

Thank you for your comments! I was directed to this post https://www.researchgate.net/post/Is_it_scientifically_accepted_to_compare_RNA-Seq_data_to_RTqPCR_data/amp It was mentioned that " Library mass normalization is performed using a different set of methods of which TMM is currently hands down the best. But when you normalize library mass - you sacrifice between gene measurements. You can compare the same gene across multiple samples - relatively, but you cannot compare two genes within the same sample". I was wondering if you have any thoughts on that? Thanks again!

[jingjin2203]

I think I understand why within-sample differential gene expression analysis is not possible after TMM because gene length normalization is skipped. Without gene length normalization, the relative quantification of the expression level of gene of interest to "internal control" gene would not be accurate. However, my goal is to compare the "relative quantification" of gene of interest across different samples. The gene length effect should cancel out across the samples, is that correct?