Entering edit mode
4.0 years ago
wenbinm
▴
40
Hi there, Is TCGA-BRCA composed of multiple independent cohorts? It was mentioned in their first publication (https://www.nature.com/articles/nature11412) that samples from 466 patients were processed on 5 different platforms. And their second publication (https://www.sciencedirect.com/science/article/pii/S0092867415011952?via%3Dihub) reported another 351 patients. Can I treat patients from the 2 publications as independent cohorts?
For example, when I am doing survival analysis, I want to have a "discovery set" and a "validation set". Using the per-existing sub-cohorts may be better than random sampling?
Thank you!
Clinical and biomedical data for TCGA-BRCA show stage, subtype, ethnicity, country of origin (and more, if you need), all of which you might want to reference before using the data from 2 publications as independent cohorts.