Hi, I am working with the Clariom D microarray, and I have annotated the PROBEID via the "clariomdhumantranscriptcluster.db" for some identifiers. However, some of the "ENSEMBL" (or other gene identifier) annotated probes does not have the corresponding transcript ID, e.g. "ENSEMBLTRANS." That sounds a bit bizarre. Is there a reason for this? Below, I'll report the code chunk I used. Thank you.

library(oligo)
library(affycoretools)
library(clariomdhumantranscriptcluster.db)
dat <- read.celfiles(list.celfiles())
eset <- rma(dat)
eset <- annotateEset(eset, clariomdhumantranscriptcluster.db, columns = c("ACCNUM","PROBEID", "ENTREZID", "SYMBOL", "GENENAME","ENSEMBL","ENSEMBLTRANS","REFSEQ","EVIDENCE"))

It's not really bizarre at all. The clariomdhumantranscriptcluster.db package really only has two tables in the underlying database that are used to do the annotations. One is called 'accessions' and one is called 'probes'. The 'probes' table maps the Affy probeset ID to NCBI Gene IDs.

So if you ask for anything more than that (like say Ensembl annotations) what happens is that the org.Hs.eg.db package is used to map from NCBI Gene ID to either Ensembl Gene ID or Ensembl Transcript ID. But that is no easy task! So what you are getting are the current mappings of NCBI Gene ID to Ensembl IDs, based on what NCBI thinks matches. It's hard enough to get the two annotation services to agree on genes, not to mention transcripts. So you should expect what you got, because mapping from NCBI to EBI/EMBL and vice versa is not trivial.