Thank you for the reply. I have another question regarding a missing gene when Affymetrix Human Genome U133 Plus 2.0 (GPL570) probes are converted to gene level expression set
I'm interested in this gene ZEB1, it has the following probes based on AnnotationDBI
select(hgu133plus2.db, keys="ZEB1", columns = c("PROBEID", "SYMBOL","GENENAME"), keytype = "SYMBOL")
SYMBOL PROBEID GENENAME
1 ZEB1 208078_s_at zinc finger E-box binding homeobox 1
2 ZEB1 210875_s_at zinc finger E-box binding homeobox 1
3 ZEB1 212758_s_at zinc finger E-box binding homeobox 1
4 ZEB1 212764_at zinc finger E-box binding homeobox 1
5 ZEB1 239952_at zinc finger E-box binding homeobox 1
If I search for "ZEB1" , it is not in the FRMA gene version of the expression set even though it should be on this array platform
If i just grep for ZEB1, i do get a read through gene ZEB1-AS1
However, I can find the probes of this gene on the probe-level expression set.
probe_id = c("210875_s_at","212758_s_at","212764_at","239952_at")
featureNames(GSE4183GPL570_RNA_ORIGINALPROBE36691) [featureNames(GSE4183GPL570_RNA_ORIGINALPROBE36691) %in% probe_id]
 "210875_s_at" "212758_s_at" "212764_at" "239952_at"
Can you explain how the ZEB1 dropped out after the probe to gene conversion step please?
Also, FRMA has different summarization method in the package (average, median, median polish, weighted average etc) , which method does inSilicoDB uses when it runs FRMA?
Thank you in advance.