Validate RNAseq Models in Microarray Data?

Question

[Vitor1]

Hi,

I built a risk score model for trying to predict melanoma outcomes in TCGA-SKCM, based in FPKM data from the patients (I know of the drawbacks of using FPKM, but this is more of a test).

I was looking at the literature and very frequently, the risk score (when built in TCGA cohort) is also validated in microarray data (https://www.aging-us.com/article/203895/text, https://www.nature.com/articles/s41598-020-69330-2, among many others).

My doubt if its really ok to do that, to validate a RNAseq model in microarray data. And if so, in what ways should I normalize the microarray data?

Thanks