how to combain microarray sets from the same platform but raw data were normalized by different methods?

Question

[lur_murad]

I want to combine three microarray datasets (E-TABM-158, GSE2034, and GSE3494) all of them from the same platform(HG-U133A). However, the raw data were normalized by RMA, MAS5, and global mean method. I want to find Fold change value for the combination set. Is it enough to use z-score before combine them and then calculate the FC value by this formula:

 if((group2_mean(k)/group1_mean(k))>=1)
         fc=(group2_mean(k)/group1_mean(k));
     else
         fc=-(group1_mean(k)/group2_mean(k));
     end

Thank you in advance

Answer 1

[pfs]

It looks like you can download the raw data for all three microarray datasets. I would take the raw data and implement the same normalization methods on each and then perform the analysis?

[lur_murad]

Yes, I have Series Matrix File as text file which contain the intensity value for each prop. Do you mean I have to download the cell files and normalize them in the same normalization method? excuse my question I am new in this area

[pfs]

Yes, I would recommend downloading the cel files, loading these files in R, then using the R program LIMMA to normalize each dataset using the same normalization method and then perform your analysis. There may be better programs than LIMMA but if I recall correctly you should be able to do the data normalization and analysis with this bioconductor package.