Some confusions on survival analysis based on TCGA prostate cancer (PRAD) dataset
chen xin
Dear Mr./Mrs.,
Recently, I read though one paper "Identification of a Novel Coregulator, SH3YL1, That Interacts With the Androgen Receptor N-Terminus". I was especially interested in the survival analysis of UBN1 based on TCGA data, which was one of the results in this paper. With the author's help, I successfully repeated their KM plot using cBioPortal for cancer genomics (http://www.cbioportal.org/).
Because they tried to divide all the patients in TCGA for prostate cancer (PRAD) based on high/low UBN1 mRNA expression. Therefore only the mRNA expression data was used for survival analysis. The detailed parameter settings were shown in attached file (the 1st slide).
In TCGA provisional cohort, there are 498 PRAD samples with mRNA expression. But in the disease free survival results (the 3rd slide), there are in total 214 samples (23 over-expressed cases and 191 other cases). And in the overall survival result (the 2nd slide), there are only 215 PRAD samples. Do you know how the other sample in PRAD TCGA data set are filtered out?
I am looking forward to your kind reply. Thanks very much!
Best Regards!
Chen Xin
JianJiong Gao
We only perform the survival analysis for the examples with survival
data. In this case, the samples with OS data and the ones with DFS
data are not identical.
Best,
-JJ
> You received this message because you are subscribed to the Google Groups
> "cBioPortal for Cancer Genomics Discussion Group" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to cbioportal+...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
chen xin
Thanks for your kind reply!
Best Regards!
Chen Xin