Using Oligo/limma for analysis of the Mouse Gene 2.0 ST array, I noticed a few un annotated  results that lack any gene annotation, upon inspecting some of these IDs on Netaffx they turned out to be the "exon" probesetids  not the transcript_ids. I noticed that for some genes both the Transcript_cluster_id and the individual probeset_ids (exon?) were present in my output. 

Looking into pd.mogene.2.0.st it does appear that this happens a times, where the transcript_cluster_id and the individual exon id are both in the "core" set. Is this an affymetrix annotation issue/feature and does anyone have a suggestion on what to do. 

con <- db(pd.mogene.2.0.st)
head(dbGetQuery(con, "select * from pmfeature inner join 
+ core_mps on pmfeature.fsetid=core_mps.fsetid where 
+ core_mps.meta_fsetid='17219448';"))
      fid   fsetid  atom    x    y meta_fsetid transcript_cluster_id   fsetid
1  216932 17219449 16055  923  134    17219448              17219448 17219449
2 1015902 17219450 16056  341  630    17219448              17219448 17219450
3 2307453 17219451 16057  680 1431    17219448              17219448 17219451
4  164045 17219452 16058 1232  101    17219448              17219448 17219452
5 1692866 17219453 16059  265 1050    17219448              17219448 17219453
6  811295 17219454 16060  458  503    17219448              17219448 17219454

For each fsetid of meta_fesetid 17219448 each of the fsetid (which I presume is an exon probeset) also appears as a core_mps

dbGetQuery(con, "select * from pmfeature inner join 
+ core_mps on pmfeature.fsetid=core_mps.fsetid where 
+ core_mps.meta_fsetid='17219449';")
     fid   fsetid  atom   x   y meta_fsetid transcript_cluster_id   fsetid
1 216932 17219449 16055 923 134    17219449              17219449 17219449

> sessionInfo()
R version 3.1.2 (2014-10-31)
Platform: x86_64-pc-linux-gnu (64-bit)

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C               LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8    LC_PAPER=en_US.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C             LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats4    parallel  stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] pd.mogene.2.0.st_2.14.1 oligo_1.30.0            Biobase_2.26.0          oligoClasses_1.28.0     RSQLite_1.0.0          
 [6] DBI_0.3.1               Biostrings_2.34.1       XVector_0.6.0           IRanges_2.0.1           S4Vectors_0.4.0        
[11] BiocGenerics_0.12.1    

loaded via a namespace (and not attached):
 [1] affxparser_1.38.0     affyio_1.34.0         BiocInstaller_1.16.1  bit_1.1-12            codetools_0.2-9       ff_2.2-13            
 [7] foreach_1.4.2         GenomeInfoDb_1.2.3    GenomicRanges_1.18.3  iterators_1.0.7       preprocessCore_1.28.0 splines_3.1.2        
[13] tools_3.1.2           zlibbioc_1.12.0      

This is one of the fun aspects of the newer generation of Affymetrix arrays; they tend to reuse the exon level probesets as controls when you summarize at the transcript level.

There does however appear to be a problem with the pd.mogene.2.0.st package:

> dbGetQuery(con, "select fsetid, type, meta_fsetid from featureSet inner join core_mps using(fsetid) where meta_fsetid in ('17219448','17219450','17219451','17219452');")
     fsetid type meta_fsetid
1  17219449    1    17219448
2  17219450    1    17219448
3  17219451    1    17219448
4  17219452    1    17219448
5  17219453    1    17219448
6  17219454    1    17219448
7  17219455    1    17219448
8  17219456    1    17219448
9  17219457    1    17219448
10 17219459    1    17219448
11 17219460    1    17219448
12 17219461    1    17219448
13 17219462    1    17219448
14 17219463    1    17219448
15 17219464    1    17219448
16 17219465    1    17219448
17 17219466    1    17219448
18 17219469    1    17219448
19 17219470    1    17219448
20 17219471    1    17219448
21 17219472    1    17219448
22 17219473    1    17219448
23 17219474    1    17219448
24 17219475    1    17219448
25 17219476    1    17219448
26 17219477    1    17219448
27 17219478    1    17219448
28 17219479    1    17219448
29 17219480    1    17219448
30 17219481    1    17219448
31 17219482    1    17219448
32 17219483    1    17219448
33 17219484    1    17219448
34 17219486    1    17219448
35 17219487    1    17219448
36 17219488    1    17219448
37 17219450    1    17219450
38 17219451    1    17219451
39 17219452    1    17219452

But the last three lines there should have a 'type' of 6, rather than 1:

> dbGetQuery(con, "select * from type_dict;")
   type                    type_id
1     1                       main
2     2              control->affx
3     3              control->chip
4     4  control->bgp->antigenomic
5     5      control->bgp->genomic
6     6             normgene->exon
7     7           normgene->intron
8     8   rescue->FLmRNA->unmapped
9     9   control->affx->bac_spike
10   10             oligo_spike_in
11   11            r1_bac_spike_at
12   12 control->affx->polya_spike
13   13        control->affx->ercc
14   14  control->affx->ercc->step

If we look at the netaffxTranscript.rda data that come with this package:

> load(paste0(path.package("pd.mogene.2.0.st"), "/extdata/netaffxTranscript.rda"))
> pd <- pData(netaffxTranscript)
> pd[pd$transcriptclusterid %in% c("17219448","17219450","17219451","17219452"),c(1:2,18)]
         transcriptclusterid probesetid       category
17219448            17219448   17219448           main
17219450            17219450   17219450 normgene->exon
17219451            17219451   17219451 normgene->exon
17219452            17219452   17219452 normgene->exon

Ideally you would be able to excise these control probesets so you don't have this stuff popping up in your top hits, but evidently the normgene->exon controls are slipping through:

> table(dbGetQuery(con, "select type from featureSet;")[,1])

     1      2      4      7      9     12
263551     18     23   5331     18     39

> table(pd[,18])

             control->affx   control->affx->bac_spike
                        18                         18
 control->affx->ercc-spike control->affx->polya_spike
                        92                         39
 control->bgp->antigenomic                       main
                        23                      33793
            normgene->exon           normgene->intron
                      1352                       5331
                  reporter                     rescue
                        82                        558
          rescue->flmrna->                       rrna
                        14                         25

There are evidently lots of control probes that are sneaking through the cracks, so pdInfoBuilder needs some updating to handle these arrays.