I have a very large imputation file (363GB in total) so I need to convert to a filevector format with GenABEL. My imputation file is in IMPUTE2 format and looks like so standard .imputed file:

   21 rs142159069    45000079   T T 1 0 0  etc
   etc

In other words, each row is a SNP with columns: chr, rsid, position, alleleA, alleleB, and then probabilities of being AA,AB or BB for each sample. There are 80 million SNPs in this file for 766 samples.

I used the following command in GenABEL to run the command:

data<-impute2databel(gen,pheno,outfile,makeprob = TRUE, old = FALSE, dataOutType = "FLOAT")

I get the following errors (and also if I remove the makeprob = TRUE, old = FALSE, dataOutType = "FLOAT" options)

Options in effect: 
     --infile    = c(21, 21, 21, 21, 21, 21, 21, 21, 21, 21)
     --outfile   = c(7, 5, 10, 6, 9, 1, 2, 8, 3, 4)
     --skiprows  = OFF
     --skipcols  = 5
     --cnrow     = ON, using data from file 'c(5,'
     --rncol     = ON, using data from file 'c(45000079,'
     --transpose = ON
     --Rmatrix   = OFF
     --nanString = c(5,
Can not open file 'c(21,' for reading
ERROR in Rstuff:failed in text2fvf_R
Index file not exists: ./tmp528216.fvi
Error in (function (cl, name, valueClass)  :

Clearly something is wrong with my impute gen file. However, I keep checking it looks normal. Has anyone had this issue before? I tried this on a much smaller file of just 10 SNPs and I also get the same error. Many thanks.