How to deal with two dataset of RNA expression

Question

[Yibin]

Hi all. I am anaylzing two dataset from GEO database. One is data of RNAseq, and under CPM and RPKM processing. Another seems from microarray,which describes the data processing as "The data were analyzed with Microarray Suite GCOS using Affymetrix default analysis settings and global scaling as normalization method". I have some questions. 1.How can I make their results comparable?

1. Is is necessary to perform another scale/normalize on the microarry data before analysis. Processing or not processing the data, the results of log Fold-change show significant differences. SOS!

[Yibin]

I use 4 gene, "HPRT1,B2M,ACTB,GAPDH" as inner reference gene. calculate the fold change in two sets, then scale by the average fold change. Dose this make sense

[dthorbur]

I would be wary about directly comparing two different techniques such as RNAseq and microrray. Even among different RNAseq datasets with different read lengths you can get quite different results.

[Yibin]

Yes

it is annoying

[biofalconch]

Other than just exploring trends of genes across different samples of the same platform, I wouldn't venture to directly compare microarrays vs RNA-seqs. But take a look at this thread to see some comments on the matter :)

[Yibin]

Thank you :)

I think I might compare the trends merely.