See the "newdata" argument in ?survfit. You can use it like this: cox.surv <- survfit(cox.fit,newdata=selected.genes) Indeed, you're going to have to build the data frame of your selected genes.
Then see ?plot.survfit for additional plotting options. That'll help you to build a "plot(cox.surv,--other arguments--)" command to draw nice survival curves. Hope this helps. Jerome On Tuesday 18 February 2003 15:04, Song, Guangchun wrote: > Content-Length: 355 > Status: R > X-Status: N > > > I'm doing prediction of the survival cases using gene expression > profiles(Affymetrix chips). Can somebody tell me how to use the Cox PH > model to select genes and make a prediction of survival? > > > Thanks. > > Guangchun > > ______________________________________________ > [EMAIL PROTECTED] mailing list > http://www.stat.math.ethz.ch/mailman/listinfo/r-help -- Jerome Asselin (J�r�me) Statistical Analyst British Columbia Centre for Excellence in HIV/AIDS St. Paul's Hospital 608 - 1081 Burrard Street Vancouver, British Columbia CANADA V6Z 1Y6 Email: [EMAIL PROTECTED] Phone: 604 806-9112 Fax: 604 806-9044 ______________________________________________ [EMAIL PROTECTED] mailing list http://www.stat.math.ethz.ch/mailman/listinfo/r-help
