How to use normalized expression values as input for limma?

Question

[Line Heylen]

Hi!

I'm a beginner in microarray analysis. I have an excel file with RMA normalized expression values: with probe IDs as rows and my RNA samples as columns. I can use this for GSEA, etc.. , but it seems that for analysis with limma in R, I would need the raw files and normalize them with affy in R. Is it possible to use my already normalized data as input for limma in R? How does my (txt) file need to look like? And can I import also a targets file, with only relevant clinical information, and not the Cy3 and Cy5 columns?

I apologize for this beginners question and hope you can help me.

Kind regards

Answer 1

[Irsan]

You just need to create an ExpressionSet object (see affy-package) from your normalized values.

eset <- ExpressionSet(yourMatrix)

It would be useful if you can add sample and feature annotation (also look in the affy-package documentation). When you have an ExpressionSet object, follow the limma users guide.

Maybe you don't have to this and you can just download your dataset in the appropriate format using GEOquery.