Hi,

Following the pdInfoBuilder vignette to produce a PDInfo package, I did the following.

However, trying to filter for 'require.entrez = TRUE' to continue with GSEA, I get an error.

I would be grateful if you could help me overcome this.

Regards,

George

ndf = list.files(pattern = ".ndf", full.names = TRUE)

xys = list.files(pattern = ".xys", full.names = TRUE)[1]

pd.100718.mm9.exp = new("NgsExpressionPDInfoPkgSeed",
ndfFile = ndf, xysFile = xys,
author = "",
email = "",
organism = "Mouse", species = "Mus musculus",
url = "")

makePdInfoPackage(pd.100718.mm9.exp, destDir = ".")

install.packages("pd.100718.mm9.exp", repos = NULL, type = "source")

library("pd.100718.mm9.exp")

xys.files = list.xysfiles(full.names = TRUE)

data = read.xysfiles(xys.files, pkgname = "pd.100718.mm9.exp")

rmadata = rma(data)

frmadata = nsFilter(rmadata, require.entrez = TRUE, require.GOBP = FALSE, require.GOCC = FALSE, require.GOMF = FALSE, var.filter = FALSE)$eset
Error in get(mapName, envir = pkgEnv, inherits = FALSE) :
  object 'pd.100718.mm9.exp_dbconn' not found

> sessionInfo()
R version 2.14.2 (2012-02-29)
Platform: x86_64-pc-mingw32/x64 (64-bit)

locale:
[1] LC_COLLATE=English_United Kingdom.1252  LC_CTYPE=English_United Kingdom.1252    LC_MONETARY=English_United Kingdom.1252 LC_NUMERIC=C                           
[5] LC_TIME=English_United Kingdom.1252    

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] BiocInstaller_1.2.1     genefilter_1.36.0       limma_3.10.3            pd.100718.mm9.exp_0.0.1 pdInfoBuilder_1.18.0    affxparser_1.26.4       RSQLite_0.11.1         
 [8] DBI_0.2-5               oligo_1.18.1            oligoClasses_1.16.0     Biobase_2.14.0         

loaded via a namespace (and not attached):
 [1] affyio_1.22.0         annotate_1.32.3       AnnotationDbi_1.16.19 Biostrings_2.22.0     bit_1.1-10            ff_2.2-11             IRanges_1.12.6        preprocessCore_1.16.0
 [9] splines_2.14.2        survival_2.36-12      tools_2.14.2          xtable_1.7-0          zlibbioc_1.0.1

You need to perform filtering manually in this case. The objects expected by nsFilter() are those created for the affy package and, for this reason, won't play nice with oligo objects at this point.

b

To further Benilton's point, the packages built by pdInfoBuilder are used by oligo to map probes to probesets, and know nothing about things like Entrez Gene IDs, etc. The package you need for that step is based on what NimbleGen says each probeset is measuring, and is built using the AnnotationForge package.

If you have a file containing the manufacturer's ID and either a column of RefSeq or GenBank (or a combination thereof) or Entrez Gene IDs, then you should be able to create an annotation package following this vignette.

If you build the package, install it, and then do

annotation(rmadata) <- <name of annotation package goes here>

then nsFilter() should work as you expect.