How to get subclass of breast tumor from RNAseq data

Question

[biopat]

Hi,

I recently got RNAseq data from some mouse mammary tumors. Now I want to determine what subtype of breast tumor it is (i.e. luminal A, luminal B, basal-like, normal-like, etc.). I saw a bunch of posts about using PAM50 for this, however, even after reading the papers it is not clear to me how to approach this problem.

I have run all the tuxedo analysis already, including cuffnorm and cuffdiff. What is the input file for using PAM50? Do I need a differential expression between normal "breast" tissue and tumor tissue as input for the PAM50 R algorithm? Also does the expression file I input need to contain only the PAM50 genes or can I use the file with all the genes in it.

Any help is appreciated!

Thank you!