I'm trying to repeat sva function using sva package following the vignettes (October 2014).

library(sva)
library(bladderbatch)
library(pamr)
library(limma)
data(bladderdata)
pheno = pData(bladderEset)
edata = exprs(bladderEset)
mod = model.matrix(~as.factor(cancer), data=pheno)
mod0 = model.matrix(~1,data=pheno)
n.sv = num.sv(edata,mod,method="leek")
n.sv
[1] 2

why i got 2 not 4?

Thanks

-Shirley

Hi Shirley,

I think you got it right. It is the same number as in the vignette here: 

http://www.bioconductor.org/packages/release/bioc/vignettes/sva/inst/doc/sva.pdf

Best, 

Jeff

OK. But I got another problem when applying to my own data.

> n.sv = num.sv(edata2,mod,method="leek")
> n.sv 
[1] 14 
> svobj = sva(edata2,mod,mod0,n.sv=n.sv) 
Number of significant surrogate variables is:  14 
Iteration (out of 5 ):Error in density.default(x, adjust = adj) : 'x' contains missing values
In addition: Warning message: 
In pf(fstats, df1 = (df1 - df0), df2 = (n - df1)) : NaNs produced
Error during wrapup: cannot open the connection  

Any suggestions?

Can I skip this step and directly run step "7 Applying the ComBat function to adjust for known batches" with this error message?

Thanks!