Association of clinical variables in Cancer Genomics or Immunology
One question we've had come up many times, whether it's in the context of cancer genomics with the TCGA dataset in OncoLand, or in the context of ImmunoLand and project-specific clinical variables, or even for your own datasets where you have many clinical parameters, is how to quickly scan all clinical variables, based on some prescribed grouping, and find the variables that are most significantly associated with that grouping.
Imagine we have a population of samples (let's say patients with Colon Adenocarcinoma) and we'd like to know - what clinical variables in the TCGA dataset correlate with that status? For instance, what clinical variables correlate with BRAF V600E mutation status in these samples?
I'm pleased to announce that we now have that ability, using the new Group Association view, available at the top level of every Land.
As you can see from the screenshot, MSI status (Microsatellite instability) correlates with BRAF V600E status in Colon Adenocarcinoma samples. A quick search of the literature finds similar conclusions: http://www.ncbi.nlm.nih.gov/pubmed/23878352.
With thousands of clinical variables available in TCGA and other datasets, this new functionality opens up the data mining possibilities for users interested in looking at clinical data side-by-side with OMIC data.