[Christopher Viot]

hi,

as already signaled  (andbelo  - 11 déc 2008 http://groups.google.com/group/rqtl-disc/browse_thread/thread/64726bb992b6d0eb#), when changing cross type (cross created with Gary format as well as with csvr format) from initial "f2" type to "riself", BBs are changed to NA ; puzzling ..

strange at it may seem, if as initial data one codes the BBs as ABs then one can have, after setting 'class(cross)[1] <- "riself"', the looked-after riself population genotype data !

[Karl Broman]

We can't handle heterozygous genotypes in RIL, and so these need to be
deleted (or treated as missing in the call to read.cross),
and the genotypes coded as if it were a backcross.

Alternatively, change all of the 3's in the genotype data to 2's (and
omit any existing 2's).

mycross <- read.cross( .... )
class(mycross)[1] <- "riself"
for(i in 1:nchr(mycross))
g <- mycross$geno[[i]]$data
g[g==2] <- NA # omit heterozygotes
g[!is.na(g) & g==3] <- 2
mycross$geno[[i]]$data <- g
}

karl

On Sep 10, 2009, at 10:44 AM, Christopher Viot wrote:

> hi,
> as already signaled (andbelo - 11 déc 2008 http://groups.google.com/group/rqtl-disc/browse_thread/thread/64726bb992b6d0eb#)

> , when changing cross type (cross created with Gary format as well