Yes, there are a few differences between these 2 files. There are 5 probes in the bgx version that are not in the txt version of the HumanRef-8_V3_0_R0_11282963 manifest. > length(intersect(fromBGX$ProbeID, fromTXT$Array_Address_Id)) [1] 24526 > setdiff(fromBGX$ProbeID, fromTXT$Array_Address_Id) [1] 2230095 290491 450670 5360291 6250095 > setdiff(fromTXT$Array_Address_Id,fromBGX$ProbeID) integer(0) Adi -----Original Message----- From: Lynn Amon [mailto:lamon@fhcrc.org] Sent: Friday, May 30, 2008 2:22 PM To: Tarca, Adi Cc: bioconductor at stat.math.ethz.ch Subject: Re: [BioC] HumanRef-8_V3 Hmmm... There seems to be a difference with the .bgx files and .txt files on the illumina.com site. In HumanRef-8_V3_0_R0_11282963_A.txt there is no probe ID 6250095 but there is such a probe on MouseRef-8_V2_0_R0_11278551_A.txt (gene = LOC654426). The probe ID for TXN in HumanRef-8_V3_0_R0_11282963_A.txt is 4290543. I don't have access to a conversion tool for the bgx files. Would anyone be willing to take a look at the differences between bgx and txt files? Lynn Tarca, Adi wrote: > Hi Lynn, > > As I said in the first e-mail, the array is HumanRef-8_V3 array. This is > an array with 8 slots for 8 different samples. > According to the HumanRef-8_V3_0_R0_11282963_A.bgx from Illumina for > this chip, this probe has "TXN" as gene symbol and is called Thioredoxin > mRNA. The accession is NM_003329.1 > It is true though that it has a missing Entrez ID. I will send you an > array so you can look at it. > > > Regards, > Adi > > > > > -----Original Message----- > From: Lynn Amon [mailto:lamon at fhcrc.org] > Sent: Friday, May 30, 2008 1:44 PM > To: Tarca, Adi > Cc: bioconductor at stat.math.ethz.ch > Subject: Re: [BioC] HumanRef-8_V3 > > Hi Adi, > That issue wasn't clear from your first email. If probe IDs are > missing, that is a bug. I don't see ProbeID = 6250095 anywhere in the > illumina annotation package for Human v3. I do find it in the Mouse v2 > array though. What accession id did you find for it? Can you send me a > > list of probe IDs that are missing. I don't have any V3 arrays to test. > Thanks, > Lynn > > > Tarca, Adi wrote: > >> Hi Lynn, >> I am aware that these packges work with probe IDs and not with target >> IDs. However there are probe IDs on the HumanRef-8_V3 that could not >> > be > >> found in either two packages lumiHumanAll.db and >> illuminaHumanv3ProbeID.db. >> >> e.g: >> >> >>> x<-illuminaHumanv3ProbeIDSYMBOL >>> aT1$Symbol<-as.vector(unlist( as.list(x[aT1$ID]) )) >>> >>> >> Error in .checkKeys(value, Lkeys(x), x at ifnotfound) : >> value for "6250095" not found >> >> However I was able to build my own annotation package using >> makeHUMANCHIP_DB function. >> >> Thanks, >> Adi >> >> >> -----Original Message----- >> From: Lynn Amon [mailto:lamon at fhcrc.org] >> Sent: Friday, May 30, 2008 12:55 PM >> To: Tarca, Adi >> Cc: bioconductor at stat.math.ethz.ch >> Subject: Re: [BioC] HumanRef-8_V3 >> >> The illumina annotation packages use the probe ID rather than the >> > target > >> id output by illumina's analysis software. If you use the .txt or >> > .csv > >> files output by the scanner, you will have probe IDs that match the >> existing annotation packages. >> Lynn Amon >> >> >> Tarca, Adi wrote: >> >> >>> Hi all, >>> >>> Did anyone make a package for the Illumina HumanRef-8_V3 array? >>> I have tried both lumiHumanAll.db and illuminaHumanv3ProbeID.db but >>> >>> >> they >> >> >>> no not recognize the IDs on this array. >>> >>> Thanks, >>> >>> Adi Tarca >>> >>> _______________________________________________ >>> Bioconductor mailing list >>> Bioconductor at stat.math.ethz.ch >>> https://stat.ethz.ch/mailman/listinfo/bioconductor >>> Search the archives: >>> >>> >> http://news.gmane.org/gmane.science.biology.informatics.conductor >> >> >>> >>> >>> >> >> > > >