List of TCGA genes differentially expressed in normal versus tumor

Question

[jamespower]

Hi,

I am looking for a gene list with differentially expressed genes in a particular tissue, using normal versus tumor samples (for example as reported in https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0071462#s3).

The TCGA website has RNA-Seq data on tumor samples, but I was just looking for the reported gene list, not to do analysis myself.

Where can I find this?

Thank you

[Kevin Blighe]

Hi James, RNA-seq / gene expression was not a primary focus of the TCGA. The data that is available at Genomics Data Commons (GDC) does not include any differential expression results for tumour versus normal. I am neither aware of any third party web-site that has such data. Your best bet is to re-analyse it yourself. I began doing this last year and it would only have taken 1 week to re-analyse the main cancers, however, I am an experienced user of TCGA data. What is stopping you from going ahead to re-analyse it? Time constraint?

[jamespower]

Thank you Kevin for your reply. Yes, time constrains, and the fact that I was hoping there were gene lists available already, since it seems like a very common and possibly simple analysis that doesn't need to be re-done by everyone...

[Kevin Blighe]

I have just done another quick search and cannot find anything. Another issue is that not many normal tissues had RNA-seq performed. I recall that some cancers had 0 normal RNA-seq samples. Breast cancer (BRCA), which is the largest cohort, has 107 normal RNA-seq samples, I believe. I think that colon or lung only had ~15 or so.

If I had the data on hand I would share it with you, but don't have it unfortunately. Firebrowse has the count matrices in RSEM and other formats: http://firebrowse.org/

RSEM can be input to DESeq2 via tximport.

Answer 1

[jamespower]

To follow-up if useful to others, I found three websites that will do this automatically:

http://gepia.cancer-pku.cn/index.html

xena.ucsc

http://www.cbioportal.org/

The gepia website seems great because it allows to do a bulk comparison, the others seem to be focused on a "by gene" comparison.

[Kevin Blighe]

Just to clarify: these web-sites that you list will perform differential expression comparisons for the user via a web GUI?

[a.james]

there is no option to download the differentially expressed genes

Answer 2

[EagleEye]

A: cancer genes database

http://ncg.kcl.ac.uk