why are RNA-seq fkpm of TCGA PAN-cancer normal and primary tumor are the same

Question

[zany1983]

I was going to compare gene expression in primary tumor and normal control.

So I downloaded PAN cancer data from xena.USCS.edu (https://tcga.xenahubs.net/download/TCGA.PANCAN.sampleMap/PANCAN_clinicalMatrix for clinical matrix and
https://toil.xenahubs.net/download/tcga_RSEM_gene_fpkm for gene expression TOIL Gene RSEM fpkm dataset.

Then I selected patient that have matched normal and primary tumor sample types, and I found that for each gene, fpm from normal and primary tumor are JUST THE SAME!

I wonder what's wrong with this.

Is there anybody helping me? Thanks in advance!

Answer 1

[Chris Miller]

In almost all cases cases, each TCGA patient only had tumor RNAseq done. (see previous discussions). Are you sure that there was actually normal RNAseq available, or have you pulled down the same file twice through different linkages?

[zany1983]

Yes, I noticed much more number of tumor than control, with many of patient having only tumor data. However, I have filtered the data from above mentioned link to make sure that each patient have both normal and primary tumor sample IDs.

I'm sure that I downloaded the intact file (https://toil.xenahubs.net/download/tcga_RSEM_gene_fpkm )only once

I'm wondering if any other person also encountered this problem? Or, should I check other portals?