Nomalization - TCGA, RNA-seq and Microarray

Question

[jain72744]

Hi, I have three different datasets- one is TCGA Data for liver cancer, then is RNA-seq data of liver cancer and then microarray data. Since the technology or chip used for each of these data is different, how can I possibly combine all three data using which type of normalization. I came across SVA package in R- but cannot figure out if it will work for all three or is there a more appropriate method?

Answer 1

[Zhenyu Zhang]

You can rank normalize them, but to tell the truth, I won't trust any conclusion you can draw from such combined data.

The recommended way is to do investigation in one dataset, and validate the finding in another.