Predicting gene expression of a cancer from the expression of subset

Question

[morovatunc]

Hi,

First of all forgive my ignorance for the idiotic headline of the question. My aim is to analyse TCGA RNA-seq information in a way that predicting a group of genes' expression.

For a single tumor sample, I know set of genes which are crucial of the cancer progression. I want to predict the other genes which are not belonged to these crucial gene group by setting up a regression model.

I would have predicted a behavior of a phenotype, say the survival or a single event with the information of the whole genes. However, I am not sure if I can do the like I asked above?

Am I doing something nonsense? Could you please turn to me to the at least the right direction?

Best,

Tunc.