My pathifier results of METABRIC (discovery + validation) and TCGA (batch corrected + normalized) are not reproducible to fully (They are reproducible to more or less 50%). My concern is whether this is due to the cross-platform variations like microarray-based (METABRIC) and RNA-Seq-based (TCGA) or something else? If yes, then why?

I don't know whether there are any biological implications governing this issue.

I would be really grateful if someone could please shed some light on this topic and share their valuable guidance.

Thanks in advance.