Using RNA-seq expression data in a classification model

Question

[knpayne2]

Hi all,

So I have noticed that for microarray data, there are countless accounts of using microarray expression data to predict some phenotypic quantity, classification, etc... I have been tasked with building a classification model using RNA-seq expression as covariates in a logistic regression, but there doesn't appear to be much literature on doing this sort of analysis. This leads to the natural question for me, is this a bad idea?

Thanks!

[Biomonika (Noolean)]

Do you already have the table with read counts ready? And do you have training set with the classes you want to to use and their expression levels? If so, it's pretty straightforward to just go ahead and see if logistic regression makes sense or not. See corresponding chapters in ISLR book: http://www-bcf.usc.edu/~gareth/ISL/