Pipeline for analyzing Microarray & RNA-seq GSE files from NCBI GEO

Question

[jsl]

0 down vote favorite I have very limited experience in R but would like to know if anyone can share their pipeline for analyzing GSE files from GEO, both for microarray and/or RNA-seq. The eventual goal would be to look at the differentially expressed genes.

For example, I would like to analyze GSE113590 which is a RNA-seq data and GSE47045 which is a microarray data.

The general consensus seems to be that you download the data using this:

source("http://bioconductor.org/biocLite.R")
biocLite("GEOquery")
library("GEOquery")
gset <- getGEO("GSE113590", GSEMatrix =TRUE)

But I'm not sure how to move forward from here, and there seems to be a different pipeline depending on whether it is a microarray / RNA-seq.

Thanks for your help.

[h.mon]

Hello junsionglow!

It appears that your post has been cross-posted to another site: https://stackoverflow.com/questions/51689293/

This is typically not recommended as it runs the risk of annoying people in both communities.

[jsl]

Understood, my apologies. I have taken down the post on stackoverflow.

Answer 1

[h.mon]

The distributions of RNAseq counts and array intensities is very different, hence the need for different packages. limma is the go-to package for microarray analysis, for RNAseq counts, the main options are edgeR, DESeq2 and limma, after using the voom transformtion on the counts.

[jsl]

I see. But I cant seem to extract the counts on the gset. I know that for microarray, one could use

gset <- getGEO("GSE47045", GSEMatrix =TRUE)
if (length(gset) > 1) idx <- grep("GPL6246", attr(gset, "names")) else idx <- 1
gset <- gset[[idx]]

str(exprs(gset))

I get

 num [1:34760, 1:24] 12.85 11.2 7.58 13.42 6.72 ...
 - attr(*, "dimnames")=List of 2
  ..$ : chr [1:34760] "10338001" "10338003" "10338004" "10338017" ...
  ..$ : chr [1:24] "GSM1143711" "GSM1143712" "GSM1143713" "GSM1143714" ..

But when it comes to RNA-seq Illumina data, the "if" command line generated NULL counts..

[h.mon]

Can you provide an example of a RNAseq accession which causes trouble?

[jsl]

this is the one im trying to analyze GSE113590

Answer 2

[ewre]

Not sure if this one is helpful for you cause it deals with ArrayExpress which is EBI instead of NCBI. As I know, most of the datasets in GEO can be found in ArrayExpress.