RNASeq relative isoform usage across batches

Question

[omicsnstuff]

I'm looking for input from experts on RNA-seq about the comparison of isoform ratio across different batches.

Assuming all things are equal when considering RNA isoform allocation of reads from RNASeq, could it be argued that the relative abundance of isoforms of a gene could be compared between different datasets or batches of patient whole blood.

We know that the relative abundance of isoforms is dependant on a variety of environmental and genomic factors. assuming the methods of detection are the same, the bioinformatics pipelines are the same, could you compare the relative abundance of transcripts from a specific gene in say 100 healthy controls from one study with perhaps two controls and one patient from another study to see if the gene has different isoform balance in the diseased patient?

I feel like this control for many technical factors (i.e. sequencing depth, gene length etc) and for the batch effect itself, by looking only at relative and not absolute abundance.

All input and comments welcome!

Answer 1

[Asaf]

Biologically, do you think that isoforms are relative or absolute? I think both apply in different cases, sometimes two isoforms are two different proteins and are regulated separately, in other cases the overall level of all isoforms is constant and the ratios differ. I think that having the analysis done on the transcript level could be beneficial but I would use it when I'm looking for a specific effect, not globally, it feels more messy than gene level.