[BioC] Kaplan Meier curve

[Heidi Dvinge]

Hi Wenhuo,

> Hi --
>
> I am trying the survival package to draw kaplan meier curve but returned
> with unexpected result.
> here is my data: d, status: 1 mean died, 0 means still alive.
>
>    gene status days
> 1    ko      1   81
> 2    ko      1  119
> 3    ko      1   92
> 4  ctrl      1   19
> 5  ctrl      1   15
> 6    ko      0   41
> 7  ctrl      1   16
> 8    ko      0   41
> 9  ctrl      1   21
> 10   ko      0   41
> 11 ctrl      0   41
> 12   ko      0  41
> 13 ctrl      1   31
>
I think there might be a slight problem with your data here, more
specifically the "days" column. The 12th data point looks like it's
formatted differently than the others (a space after it). That might give
rise to the one alive mouse showing up as a cross at ~day 100.

Incidentally, something like this looks right:

d <- data.frame(gene=c(1,1,1,2,2,1,2,1,2,1,2,1,2),
status=c(1,1,1,1,1,0,1,0,1,0,0,0,1),
days=c(81,119,92,19,15,41,16,41,21,41,41,41,31))
survfit(Surv(days, status)~gene, data=d) -> fit
plot(fit, conf.int=F, col=2:3, lty=1:2)

That gives 8 deaths in total, and 5 alive mice (plotted on top of each
other at day41).

HTH
\Heidi


> survfit(Surv(days, status)~gene, data=d) -> fit
> plot(fit, conf.int=F, col=2:3, lty=1:2)
>
> as attached, the green one is ko genotype which can live longer. But there
> should have alive mice left, 4 out of 7, as the data shows. Could any help
> me on this point?
>
>
>
> Wenhuo Hu
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at r-project.org
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives:
> http://news.gmane.org/gmane.science.biology.informatics.conductor