Which method could be used for cross normailizing microarray samples from different platforms namely (Affmetrix, Agilent,Illumina..) ?

Question

[macmath]

Received data from Affmetrix

New data in Agilent and other in Illumina

How to normalize these data so that we can reduce platform batch effect

Answer 1

[Manvendra Singh]

There is an algorithm from ComBat which is here also in sva package from bioconductor here.

When I analyze I just pick single value for each gene on their mean, I merge normalized data from different platforms . ( e.g. "data")

I normalize merged data on their quantiles

by

library(limma)
normalized.data <- normalizeQuantiles(data)

hth

[macmath]

Thank you for your suggestion Manvendra Singh

[macmath]

This package has problems in regard to bias in expression values. Are you aware of some other tools?

[Manvendra Singh]

Which tool?

ComBat is good for expression bias