Hi Lynn and Marc, Thanks for your help and trying to get to the source of the problem. Perhaps BMP4, 1940386 is an anomaly. regards julian ----- Original Message ----- From: "Lynn Amon" <lamon@fhcrc.org> To: "Marc Carlson" <mcarlson at="" fhcrc.org=""> Cc: "Julian Lee" <julian at="" omniarray.com="">, "bioconductor" <bioconductor at="" stat.math.ethz.ch=""> Sent: Monday, November 3, 2008 10:41:14 AM GMT -08:00 US/Canada Pacific Subject: Re: [BioC] BMP4 probe '1940386' missing in illuminaHumanv2BeadID.db (as compared to illuminaHumanv2ProbeID.db) Julian, I no longer maintain the Illumina annotation packages but I can tell you that I made the illuminaHumanv2ProbeID.db package using annotation provided by Nuno Barbosa-Morais who BLASTed the probe sequences against hg18. I only used probes which had 100% similarity with refseq sequences. You can find the BLAST results at: http://www.compbio.group.cam.ac.uk/Resources/Annotation/ >From what I can see by BLASTing the three probe sequences again against the current build, all three have perfect matches with refseq sequences for BMP4 1940386 AGTAGAGGGATGTGGGTGCCGCTGAGATCAGGCAGTCCTTGAGGATAGAC NM_130851.2, NM_138050.2, NM_001202.3 3360703 GAGACGCAGACGCAGAGGTCGAGCGCAGGCCGAAAGCTGTTCACCGTTTT NM_130851.2 1850561 ACGCCGCTGCTGCTCCGGCTGAGTATCTAGCTTGTCTCCCCGATGGGATT NM_001202.3 Also, all three probes are listed with in the last available Illumina-provided annotation file that I can find: HumanWG-6_V2_0_R2_11223189_A 1940386 NM_130851.1 3360703 NM_130851.1 1850561 NM_001202.2 So, I'm a little confused as to why 1940386 is not showing BMP4 as the symbol name in the new annotation package. What are the other annotation results for this probe? RefSeq? Accession? Marc, do you know what annotation file was used for these packages? Lynn Amon Marc Carlson wrote: > Hi Julian, > > We are not responsible for the mappings used to tie the illumina IDs > onto the gene IDs, for those you probably need to talk to the > manufacturers. When we build a new package, all we do is connect those > manufacturer provided mappings to the data from public repositories. If > the manufacturer changes their mind about how one of those probes should > map we have little choice but to believe them. But without even seeing > these new manufacturer mappings, I would guess that there is probably > nothing wrong with this package. It is (unfortunately) fairly common > for manufacturers of microarrays to decide that a probe does not really > measure what they originally thought it measured. This is part of why > we rebuild all these packages every six months. We are trying our best > to give you the most current/accurate picture possible. > > If you have doubts about the correct mapping of that probe, then please > check with the manufacturer to see what they claim their platform > measures. If there are any discrepancies in the package from this, then > please let me know immediately. > > > Marc > > > > > Julian Lee wrote: > >> Hi, >> >> may i know the differences between the 2 packages and why is this so. I have a gene of interest on the Illumina platform and have been using the IlluminaHumanv2ProbeID.db package to annotate it, but things have changed when i moved to R.2.8.0 on illuminaHumanv2BeadID.db package. >> >> I'll illustrate it by an example >> >> Gene of interest - BMP4 >> >> R.2.7.1 >> illuminaHumanv2ProbeID.db package >> >> >> >>> bmp4_p<-as.character(unlist(mget('BMP4',revmap(illuminaHumanv2Prob eIDSYMBOL)))) >>> bmp4_p >>> >>> >> [1] "1850561" "1940386" "3360703" >> >> >> >>> sessionInfo() >>> >>> >> R version 2.7.1 (2008-06-23) >> i386-pc-mingw32 >> >> locale: >> LC_COLLATE=English_United States.1252;LC_CTYPE=English_United States.1252;LC_MONETARY=English_United States.1252;LC_NUMERIC=C;LC_TIME=English_United States.1252 >> >> attached base packages: >> [1] tools stats graphics grDevices utils datasets methods >> [8] base >> >> other attached packages: >> [1] illuminaHumanv2ProbeID.db_1.1.1 AnnotationDbi_1.2.2 >> [3] RSQLite_0.6-9 DBI_0.2-4 >> >> All looks well. However, when i upgraded to R.2.8.0, and installed the illuminaHumanv2BeadID.db package, i encountered these results >> >> R.2.8.0 >> illuminaHumanv2BeadID.db package >> >> >> >>> bmp4_p<-as.character(unlist(mget('BMP4',revmap(illuminaHumanv2Bead IDSYMBOL)))) >>> bmp4_p >>> >>> >> [1] "1850561" "3360703" >> >> >> >>> sessionInfo() >>> >>> >> R version 2.8.0 (2008-10-20) >> i386-pc-mingw32 >> >> locale: >> LC_COLLATE=English_United States.1252;LC_CTYPE=English_United States.1252;LC_MONETARY=English_United States.1252;LC_NUMERIC=C;LC_TIME=English_United States.1252 >> >> attached base packages: >> [1] tools stats graphics grDevices utils datasets methods >> [8] base >> >> other attached packages: >> [1] illuminaHumanv2BeadID.db_1.1.2 RSQLite_0.7-1 >> [3] DBI_0.2-4 AnnotationDbi_1.4.0 >> [5] Biobase_2.2.0 >> >> >> The probeID/beadID "1940386" has disappeared. Why is this so? is there a mistake in the illuminaHumanv2BeadID package? >> Is it possible to achieve reproducible results by upgrading to 2.8.0? >> >> many thanks >> >> julian >> >> >> >> > > _______________________________________________ > 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 > -- Julian Lee Bioinformatics Specialist Cellular and Molecular Research National Cancer Center Singapore