Dear List, The following question was asked before before but I do not believe that an answer was given: Should Illumina methylation beta data be transforrmed before input into Limma. Beta values are between 0 and 1. In ordinary frequentist statistics variables which range between 0 and 1 are subject to a logistic transformation before anova is applied to them (more precisey a general linear model is used with a logistic link function and a binomial error model). Should A. A logistic transformation be applied before inputing methylation beta data into Limma. or B. Should a log transformation be applied. or C. Should some other transformation be applied. or D, Should no transformation be applied. I would greatly appreciate any insight that you might have into this problem. Thanks and bet wishes, Rich ------------------------------------------------------------ Richard A. Friedman, PhD Associate Research Scientist, Biomedical Informatics Shared Resource Herbert Irving Comprehensive Cancer Center (HICCC) Lecturer, Department of Biomedical Informatics (DBMI) Educational Coordinator, Center for Computational Biology and Bioinformatics (C2B2)/ National Center for Multiscale Analysis of Genomic Networks (MAGNet) Room 824 Irving Cancer Research Center Columbia University 1130 St. Nicholas Ave New York, NY 10032 (212)851-4765 (voice) friedman at cancercenter.columbia.edu http://cancercenter.columbia.edu/~friedman/ "Did he win the Nobel prize or the Ig Nobel prize for levitating the frog?". Rose Friedman, age 14

On Mon, Jan 31, 2011 at 9:57 AM, Richard Friedman < friedman@cancercenter.columbia.edu> wrote: > Dear List, > > The following question was asked before before but I do not believe > that an answer was given: > Should Illumina methylation beta data be transforrmed before input into > Limma. > Beta values are between 0 and 1. > In ordinary frequentist statistics variables which range between 0 and 1 > are subject to a logistic > transformation before anova is applied to them (more precisey a general > linear model is used with > a logistic link function and a binomial error model). > Should > A. A logistic transformation be applied before inputing methylation beta > data into Limma. > or > B. Should a log transformation be applied. > or > C. Should some other transformation be applied. > or > D, Should no transformation be applied. > > I would greatly appreciate any insight that you might have into this > problem. > > Hi, Richard. Have a look at this publication which recommends forming ratios: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3012676 Sean [[alternative HTML version deleted]]