survival curves for Leonid

January 7, 2011
By

(This article was first published on rforcancer, and kindly contributed to R-bloggers)

Leonid asked me to do a quick survival analysis of two different types of mouse (m430 and m210) with surgically implanted tumours (or something like that). The data was in the wrong format but after transforming it looked like this:

In my opinion there are far too few samples to detect any significant difference between the mouse types in anything but the most extreme case. Nevertheless it is useful for him to go through the motions and publish what he has....or rather what I do for him.

I used the survival package for R . Here is the script and results.

##attach column names to the search path
> attach(leoS)
 
## make a survival object
> leoS.surv= Surv(time,delta)
 
##fit the data separating by mouse type
> my.fit= survfit(leoS.surv~type)
 
## use the G-rho tests for difference between the curves
> sdf= survdiff(leoS.surv~type)
 
## survdiff( ) is described here: Harrington, D. P. and Fleming, T. R. (1982). A class of rank test procedures for censored survival data.
 
>sdf
Call:
survdiff(formula = leoS.surv ~ type)
 
          N Observed Expected (O-E)^2/E (O-E)^2/V
type=m210 8        7     7.12   0.00208   0.00543
type=m430 8        6     5.88   0.00252   0.00543
 
 Chisq= 0  on 1 degrees of freedom, p= 0.941 


## and now plot it all together


> plot(my.fit, lty=1:2, mark.time=T)
> legend(10, .2, c("m210", "m430"), lty=1:2)
> title(main='survdiff(formula = leoS.surv ~ type)', sub='Chisq= 0  on 1 degrees of freedom, p= 0.941', ylab="n= 8 in each group", xlab="days")
 
Which gives....
 
which I think is good enough for such a small experiment. I saved the data, images and RData in Desktop/Projects/Leonid/mouse_survival.

 

To leave a comment for the author, please follow the link and comment on his blog: rforcancer.

R-bloggers.com offers daily e-mail updates about R news and tutorials on topics such as: visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...



If you got this far, why not subscribe for updates from the site? Choose your flavor: e-mail, twitter, RSS, or facebook...

Comments are closed.