I am currently building models for predicting microRNA targets using microArray gene expression data. The predictions would be more reliable if NGS expression datasets can be used.
I agree with Luciano you can use TCGA's data. In particular you can try the function TCGAquery_integrate to have the idea of the common (matched) samples among platforms for a selected tumor, and then download and prepare the data. In addition it should be nice to use TCGAbiolinks with your recent miRLAB Bioconductor package.
For further informations or collaborations you can write me ([email protected]).
Best,
Antonio Colaprico
Article TCGAbiolinks: An R/Bioconductor package for integrative anal...