[BioC] exonmap/xmapcore and splicing.index

[Peter Saffrey]

I am analysing Affymetrix mouse exon arrays (MoEx-1_0) using xmapcore 
and would like to use the splicing index function to compare the levels 
of splicing in two conditions. It's documented here:

http://svitsrv25.epfl.ch/R-doc/library/exonmap/html/si.html

Unfortunately, this function is only available in exonmap, the (now 
deprecated) version of xmapcore. I have managed to install exonmap but I 
get errors calling si. Here is what I do:

# start exonmap and connect to db
library(exonmap)
xmapConnect()
# read array data and compute rma
adata <- read.exon()
adata at cdfName <- "mouseexonpmcdf"
x.rma <- rma(adata)
# setup the appropriate columns for the two conditions
gps <- list(c("KC4MoEX1E141.CEL", "KC4MoEX1E142.CEL", 
"KC4MoEX1E143.CEL"), c("NB1MoEX1H1.CEL", "NB1MoEX1H2.CEL", 
"NB1MoEX1H3.CEL"))
# call splicing index algorithm
mysi <- si(x.rma, c("ENSMUSG00000000001", "ENSMUSG00000000003"), gps=gps)

and the error:

Error in fix.by(by.y, y) : 'by' must specify valid column(s)
In addition: Warning message:
In is.na(cols) : is.na() applied to non-(list or vector) of type 'NULL'

Stepping down through the functions it's calling, it looks like a 
problem in gene.to.exon.probeset, which is not working and returning an 
empty list.

I thought about trying to port this function to xmapcore, but I'm not 
very experienced in R and I don't even know how to get at the source for 
the modules I've downloaded.

Is anybody using xmapcore/exonmap to do alternative splicing analysis? 
Can they provide me with any pointers or alternative tools?

Peter