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
•
written
by
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
The gepia website seems great because it allows to do a bulk comparison, the others seem to be focused on a "by gene" comparison.
Just to clarify: these web-sites that you list will perform differential expression comparisons for the user via a web GUI?
Log in to answer this question.
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?
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...
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.