Hello everybody, I'm currently working with Methylation data. Three data types: IlluminaDNAMethylation_OMA003_CPI_CPI, IlluminaDNAMethylation_OMA002_CPI_CPI both from Illumina Golden Gate BeadArray platform and Illumina Infinium Human DNA Methylation 27 platform After several trials, I finally succeeded loading the data into R(after reorganizing it into one single file - like in the example- data from the methylumi-package in R) But how do I get the mappings of the probe, e.g. Composite Element Cy3 Cy5 Detection Pvalue cg08367223 9.75030715328155 51.3125 3.68E-38 Interpretation: Cy5/Cy3 > 0 --> location x is hypermethylated, But where is this location x? Are there some mappings, similar to cdf- files from Affy-data, or is the location somehow encoded in the Composite element. The OMA003-platform for instance has comosites such as: SOX17_P303_F this means, it refers somehow to the SOX17 - gene, P303 - position 303 starting from startcodon? F? would be very greatfull for some tipps. thank's, Noemi

On Wed, Jun 23, 2010 at 6:05 PM, Noemi Andor <noemi.andor@campus.lmu.de>wrote: > Hello everybody, > > I'm currently working with Methylation data. Three data types: > IlluminaDNAMethylation_OMA003_CPI_CPI, > IlluminaDNAMethylation_OMA002_CPI_CPI > > both from Illumina Golden Gate BeadArray platform > > and > > Illumina Infinium Human DNA Methylation 27 platform > > After several trials, I finally succeeded loading the data into R(after > reorganizing it into one single file - like in the example-data from the > methylumi-package in R) > > But how do I get the mappings of the probe, e.g. > > Composite Element Cy3 Cy5 Detection Pvalue > cg08367223 9.75030715328155 51.3125 3.68E-38 > > Interpretation: Cy5/Cy3 > 0 --> location x is hypermethylated, > > But where is this location x? Are there some mappings, similar to cdf-files > from Affy-data, or is the location somehow encoded in the Composite element. > > The OMA003-platform for instance has comosites such as: > > SOX17_P303_F > > this means, it refers somehow to the SOX17 - gene, P303 - position 303 > starting from startcodon? F? > > would be very greatfull for some tipps. > > For the Illumina 27k methylation platform, see: http://bioconductor.org/packages/release/data/annotation/html/Illumina HumanMethylation27k.db.html For the other two arrays, those appear to be custom OMAs, so you will want to export the data from GenomeStudio including as much annotation as possible. Those annotation columns will end up as featureData in the resulting MethyLumiSet object. Sean [[alternative HTML version deleted]]

On Thu, Jun 24, 2010 at 1:22 AM, Noemi Andor <noemi.andor@campus.lmu.de>wrote: > thank you! > > Is GenomeStudio free available? I see on the website, that one must > register and they are asking for some costumer number. > > I do not believe so. You may need to contact your collaborator about the annotation. Sean > > > Ursprüngliche Nachricht: > > Von: Sean Davis <sdavis2@mail.nih.gov> > > An: Noemi Andor <noemi.andor@campus.lmu.de> > > Kopie: bioconductor@stat.math.ethz.ch > > Datum: Thu Jun 24 00:56:46 CEST 2010 > > > > On Wed, Jun 23, 2010 at 6:05 PM, Noemi Andor <noemi.andor@campus.lmu.de> >wrote: > > > > > Hello everybody, > > > > > > I'm currently working with Methylation data. Three data types: > > > IlluminaDNAMethylation_OMA003_CPI_CPI, > > > IlluminaDNAMethylation_OMA002_CPI_CPI > > > > > > both from Illumina Golden Gate BeadArray platform > > > > > > and > > > > > > Illumina Infinium Human DNA Methylation 27 platform > > > > > > After several trials, I finally succeeded loading the data into R(after > > > reorganizing it into one single file - like in the example-data from > the > > > methylumi-package in R) > > > > > > But how do I get the mappings of the probe, e.g. > > > > > > Composite Element Cy3 Cy5 Detection Pvalue > > > cg08367223 9.75030715328155 51.3125 3.68E-38 > > > > > > Interpretation: Cy5/Cy3 > 0 --> location x is hypermethylated, > > > > > > But where is this location x? Are there some mappings, similar to > cdf-files > > > from Affy-data, or is the location somehow encoded in the Composite > element. > > > > > > The OMA003-platform for instance has comosites such as: > > > > > > SOX17_P303_F > > > > > > this means, it refers somehow to the SOX17 - gene, P303 - position 303 > > > starting from startcodon? F? > > > > > > would be very greatfull for some tipps. > > > > > > > > For the Illumina 27k methylation platform, see: > > > > > http://bioconductor.org/packages/release/data/annotation/html/Illumi naHumanMethylation27k.db.ht > > ml > > > > For the other two arrays, those appear to be custom OMAs, so you will > want > > to export the data from GenomeStudio including as much annotation as > > possible. Those annotation columns will end up as featureData in the > > resulting MethyLumiSet object. > > > > Sean > > > [[alternative HTML version deleted]]

On Thu, Jun 24, 2010 at 12:45 PM, Noe Andor <noemi.andor@yahoo.com> wrote: > So maybe that's why I had that trouble loading the data from R, becouse > it's not from beadStudio. Methylumi is designed for data comming from > beadstudio, isn't it? Are there also some packages available for other > methylation data? MethylumiR works only with header like: > > TargetID ProbeID xx.AVG_Beta xx.Signal CY3 xx.Signal CY5 > xx.Detection Pval > > And my data looks either like: > > Hybridization REF xx xx xx > Composite Element REF Cy3 Cy5 Detection Pvalue > > or like: > > Hybridization REF xx xx xx xx xx xx xx > Composite Element REF Methylated_Signal_Intensity (M) > Un-Methylated_Signal_Intensity (U) Negative_Control_Grn_Avg_Intensity > Negative_Control_Grn_STDERR Negative_Control_Red_Avg_Intensity > Negative_Control_Red_STDERR Detection_P_Value > > (xx being the sample ID) > > Now what I did for the first data-type is adjusting the header for the > methylumi, that's kind of intuitive, and more similar to the methylumi > header than the second one. I'm not even shure with this data if I did well, > as I introduced a increasing probe-id and a beta value of 0 everywhere - > that coukd affect further normalization proceedures, doesn't it? > As for the other data-type - I could replace unmethylated with Cy3 and > methylated with Cy5, but that doesn't seem right, just to keep replacing, > deleting and adding column-names - I have no other ideas what to do. > I also tried to adjust that so called substitutions-file within the > methylumi package, to contain the columnNames of the second type as well - > this didn't work either. I'm running out of ideas... > These are just tab-delimited text files? If so, then you should be able to just use read.table() to get the data into R. After that, you can construct an ExpressionSet or MethyLumiSet as you see fit. Unfortunately, I don't think this is going to be a task that can be solved easily without some basic R programming. Sean > ------------------------------ > *Von:* Sean Davis <sdavis2@mail.nih.gov> > *An:* Noemi Andor <noemi.andor@campus.lmu.de> > *CC:* Bioconductor Newsgroup <bioconductor@stat.math.ethz.ch> > *Gesendet:* Donnerstag, den 24. Juni 2010, 2:38:35 Uhr > *Betreff:* Re: Re: [BioC] methylumi > > > > On Thu, Jun 24, 2010 at 1:22 AM, Noemi Andor <noemi.andor@campus.lmu.de>wrote: > >> thank you! >> >> Is GenomeStudio free available? I see on the website, that one must >> register and they are asking for some costumer number. >> >> > I do not believe so. You may need to contact your collaborator about the > annotation. > > Sean > > >> >> > Ursprüngliche Nachricht: >> > Von: Sean Davis <sdavis2@mail.nih.gov> >> > An: Noemi Andor <noemi.andor@campus.lmu.de> >> > Kopie: bioconductor@stat.math.ethz.ch >> > Datum: Thu Jun 24 00:56:46 CEST 2010 >> > >> > On Wed, Jun 23, 2010 at 6:05 PM, Noemi Andor <noemi.andor@campus.lmu.de>> >wrote: >> > >> > > Hello everybody, >> > > >> > > I'm currently working with Methylation data. Three data types: >> > > IlluminaDNAMethylation_OMA003_CPI_CPI, >> > > IlluminaDNAMethylation_OMA002_CPI_CPI >> > > >> > > both from Illumina Golden Gate BeadArray platform >> > > >> > > and >> > > >> > > Illumina Infinium Human DNA Methylation 27 platform >> > > >> > > After several trials, I finally succeeded loading the data into >> R(after >> > > reorganizing it into one single file - like in the example-data from >> the >> > > methylumi-package in R) >> > > >> > > But how do I get the mappings of the probe, e.g. >> > > >> > > Composite Element Cy3 Cy5 Detection Pvalue >> > > cg08367223 9.75030715328155 51.3125 3.68E-38 >> > > >> > > Interpretation: Cy5/Cy3 > 0 --> location x is hypermethylated, >> > > >> > > But where is this location x? Are there some mappings, similar to >> cdf-files >> > > from Affy-data, or is the location somehow encoded in the Composite >> element. >> > > >> > > The OMA003-platform for instance has comosites such as: >> > > >> > > SOX17_P303_F >> > > >> > > this means, it refers somehow to the SOX17 - gene, P303 - position 303 >> > > starting from startcodon? F? >> > > >> > > would be very greatfull for some tipps. >> > > >> > > >> > For the Illumina 27k methylation platform, see: >> > >> > >> http://bioconductor.org/packages/release/data/annotation/html/Illum inaHumanMethylation27k.db.ht >> > ml >> > >> > For the other two arrays, those appear to be custom OMAs, so you will >> want >> > to export the data from GenomeStudio including as much annotation as >> > possible. Those annotation columns will end up as featureData in the >> > resulting MethyLumiSet object. >> > >> > Sean >> > >> > > > [[alternative HTML version deleted]]