I have gene-expression data of 80 cancer patients coupled with survival data for the same patients. What is the most robust way to calculate and plot survival curves for all my genes using the survival data?
Basically I would like to find those genes whose expression significantly affects survival.
The survival package in R is what I have used in the past. This site has a good overview of plotting the curves and this site has more detail.
I have only done that analysis with WT vs mutant mice, i.e. 1 gene in approx states. Unless you have some extreme effects in particular genes you may struggle to get the statistical power needed if you are examining multiple genes in multiple states over a fairly small cohort
These are miRNA expression data and it seems to be some effects allready.
How did you determine the cutoff value for the expression data?