Affymetrix miRNA array normalization

0

Entering edit mode

Kamila Naxerova ▴ 100

@kamila-naxerova-4164

Last seen 10.3 years ago

Dear list, I was searching the archives for discussions about the Affymetrix miRNA chip... there are only a few threads and most of them are almost a year old. I was wondering whether in the meantime people had gathered some experience with analysis strategies - normalization in particular - and were willing to share some insights. I do not feel comfortable using RMA because I doubt that total miRNA expression can be assumed to be constant across samples (particularly when studying developmental processes or cancer). Has anybody played with the options? Thanks! Kamila

miRNA Normalization miRNA Normalization • 1.5k views

ADD COMMENT • link updated 14.5 years ago by Thomas Hampton ▴ 20 • written 14.5 years ago by Kamila Naxerova ▴ 100

0

Entering edit mode

Thomas Hampton ▴ 20

@thomas-hampton-3936

Last seen 10.3 years ago

Great question. At some point, quantile normalization would not be expected to work, because the true distribution of expression values under different conditions would just be too different. You'd see "differential expression" that was an artifact of normalization. I have seen RMA cause that with defective arrays. While I am no expert on all the arguments, I think the proof of the pudding is in the verification of miRNAs your statistical approach found differentially expressed. Best, Tom On Jul 10, 2010, at 11:26 AM, Kamila Naxerova wrote: > Dear list, > > I was searching the archives for discussions about the Affymetrix > miRNA chip... there are only a few threads and most of them are > almost a year old. I was wondering whether in the meantime people > had gathered some experience with analysis strategies - > normalization in particular - and were willing to share some > insights. I do not feel comfortable using RMA because I doubt that > total miRNA expression can be assumed to be constant across samples > (particularly when studying developmental processes or cancer). Has > anybody played with the options? > > Thanks! > Kamila > > _______________________________________________ > 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

ADD COMMENT • link 14.5 years ago Thomas Hampton ▴ 20

0

Entering edit mode

Steve Shen ▴ 330

@steve-shen-3743

Last seen 10.3 years ago

Will affymetrix qc tool wook? http://www.affymetrix.com/estore/partners_programs/programs/developer/ tools/devnettools.affx#miRNAQC steve On Sat, Jul 10, 2010 at 11:26 AM, Kamila Naxerova <naxerova@fas.harvard.edu>wrote: > Dear list, > > I was searching the archives for discussions about the Affymetrix miRNA > chip... there are only a few threads and most of them are almost a year > old. I was wondering whether in the meantime people had gathered some > experience with analysis strategies - normalization in particular - and were > willing to share some insights. I do not feel comfortable using RMA because > I doubt that total miRNA expression can be assumed to be constant across > samples (particularly when studying developmental processes or cancer). Has > anybody played with the options? > > Thanks! > Kamila > > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: > http://news.gmane.org/gmane.science.biology.informatics.conductor > [[alternative HTML version deleted]]

ADD COMMENT • link 14.5 years ago Steve Shen ▴ 330

0

Entering edit mode

Kamila vsnrma (in the vsn package) might be an option for you to try that provides the same probe summary as "normal" rma, but a less aggressive between-array normalisation (namely, only a shifting and scaling, i.e. affine-linear). See also Deepayan Sarkar et al. in NAR 2009, "Quality Assessment and Data Analysis for microRNA Expression Arrays", http://nar.oxfordjournals.org/cgi/content/full/37/2/e17 In addition, if you have "negative controls", i.e. probes on the array for which you are willing to believe that they most of them don't change expression, then you can fit the shifting and scaling factors on these, and then apply to all probes. Best wishes Wolfgang On Jul/11/10 5:11 PM, Steve Shen wrote: > Will affymetrix qc tool wook? > http://www.affymetrix.com/estore/partners_programs/programs/develope r/tools/devnettools.affx#miRNAQC > > steve > > On Sat, Jul 10, 2010 at 11:26 AM, Kamila Naxerova > <naxerova at="" fas.harvard.edu="">wrote: > >> Dear list, >> >> I was searching the archives for discussions about the Affymetrix miRNA >> chip... there are only a few threads and most of them are almost a year >> old. I was wondering whether in the meantime people had gathered some >> experience with analysis strategies - normalization in particular - and were >> willing to share some insights. I do not feel comfortable using RMA because >> I doubt that total miRNA expression can be assumed to be constant across >> samples (particularly when studying developmental processes or cancer). Has >> anybody played with the options? >> >> Thanks! >> Kamila >> >> >> _______________________________________________ >> 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 >> > > [[alternative HTML version deleted]] > > _______________________________________________ > 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 -- Wolfgang Huber EMBL http://www.embl.de/research/units/genome_biology/huber

ADD REPLY • link 14.4 years ago Wolfgang Huber ★ 13k

0

Entering edit mode

Di Wu ▴ 190

@di-wu-1837

Last seen 10.3 years ago

normalizeCyclicLoess function in limma may help. But I don't know whether it is publicly available now. It will be in new version of Bioconductor. See, http://bioconductor.org/packages/2.6/bioc/manuals/limma/man/limma.pdf More array weights should be given to the control probes as you may consider they are constant across arrays. As the author siad normalise.loess may be another option. Cheers, Di On Sun, Jul 11, 2010 at 1:26 AM, Kamila Naxerova <naxerova@fas.harvard.edu>wrote: > Dear list, > > I was searching the archives for discussions about the Affymetrix miRNA > chip... there are only a few threads and most of them are almost a year > old. I was wondering whether in the meantime people had gathered some > experience with analysis strategies - normalization in particular - and were > willing to share some insights. I do not feel comfortable using RMA because > I doubt that total miRNA expression can be assumed to be constant across > samples (particularly when studying developmental processes or cancer). Has > anybody played with the options? > > Thanks! > Kamila > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: > http://news.gmane.org/gmane.science.biology.informatics.conductor > [[alternative HTML version deleted]]

ADD COMMENT • link 14.5 years ago Di Wu ▴ 190

Login before adding your answer.