Question: RAW data in matrix: RMA normalization?
0
gravatar for alexvpickering
3.7 years ago by
alexvpickering110 wrote:

I'm doing some analysis on GSE7829. The data supplied in the GSE Matrix has been processed by Z-normalization (instead of the more common RMA). The raw intensity values are also available  in the GSE SOFT "RAW" column and I would like to RMA-normalize these raw values. How do I approach this (the CEL files are not supplied)?

I have extracted the "RAW" column from the GSE SOFT file and built an ExpressionSet object. I'm stuck here as the RMA functions take either CEL or AffyBatch objects as input. Any suggestions?

affy rma raw expression • 1.3k views
ADD COMMENTlink modified 3.7 years ago by James W. MacDonald51k • written 3.7 years ago by alexvpickering110
Answer: RAW data in matrix: RMA normalization?
2
gravatar for James W. MacDonald
3.7 years ago by
United States
James W. MacDonald51k wrote:

The data for that experiment are spotted cDNA arrays, not Affy arrays, so the RMA algorithm does not apply.

ADD COMMENTlink written 3.7 years ago by James W. MacDonald51k

Is there an appropriate normalization algorithm for cDNA arrays that will allow me to compare samples (and that keeps values as intensity measures)? I am using the characteristic direction approach to differential expression and was advised by the developers to use intensity measures (as opposed to log2 transformation, or I suppose in this case standard normal transformation).

ADD REPLYlink written 3.7 years ago by alexvpickering110

There are any number of normalization methods for cDNA arrays, and what I might think is appropriate may well be quite different than what you want. Your best bet is to look at the limma User's Guide - limma is probably the most widely used tool for dealing with cDNA arrays - and decide what normalization seems most reasonable. I have no idea what ''characteristic direction approach" means, so I can't help you there, but I believe you can tell limma not to take logs.

ADD REPLYlink written 3.7 years ago by James W. MacDonald51k
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 136 users visited in the last hour