Hi, I've generated several MA/RI plots after using RMA and noticed that the standard deviation in M (fold change) is related to A (average intensity) - to derive the SD in M, only sample fold changes over a limited range in A (a similar calculation to the Quackenbush sliding Z-score). When using RMA, the standard deviation in M is proportional to A, except for the largest intensities. However, when using expresso with all the options set to MAS, the standard deviation is independent of A. I believe the variance in M should be independent of A, in which case this suggests MAS is the way to go. Has anybody else seen this effect, or can suggest a remedy, or more likely a gap in my logic? Best wishes and thanks in advance, Harry -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~ Dr Andrew Harrison Tel: 44 (0) 207 679 3890 Biomolecular Structure and Modelling Unit Fax: 44 (0) 207 679 7193 Biochemistry and Molecular Biology Dept. University College London Gower Street Email: harry@biochem.ucl.ac.uk London, WC1E 6BT, UK http://www.biochem.ucl.ac.uk/~harry ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~

Dear Andrew, This is very often the case, I have found that Wolfgang Huber's vsn package does a very good job of taking out the variance's dependence on the intensity especially in the lower ranges. Best Susan Holmes Statistics On Thu, 22 May 2003, Andrew Harrison wrote: > Hi, > > I've generated several MA/RI plots after using RMA and > noticed that the standard deviation in M (fold change) is related > to A (average intensity) - to derive the SD in M, only sample fold > changes over a limited range in A (a similar calculation to the > Quackenbush sliding Z-score). > > When using RMA, the standard deviation in M is proportional > to A, except for the largest intensities. However, when using > expresso with all the options set to MAS, the standard deviation is > independent of A. I believe the variance in M should be independent > of A, in which case this suggests MAS is the way to go. > > Has anybody else seen this effect, or can suggest a remedy, > or more likely a gap in my logic? > > Best wishes and thanks in advance, > Harry > > -- > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~ > Dr Andrew Harrison Tel: 44 (0) 207 679 3890 > Biomolecular Structure and Modelling Unit Fax: 44 (0) 207 679 7193 > Biochemistry and Molecular Biology Dept. > University College London > Gower Street Email: harry@biochem.ucl.ac.uk > London, WC1E 6BT, UK http://www.biochem.ucl.ac.uk/~harry > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~ > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor >

ive oberved exactly the oposite. SDs for RMA,dCHip and MAS 5.0 depend on A. but MAS 5.0 is by way more dependent. look at http://www.biostat.jhsph.edu/~ririzarr/sd.html for some pictures. keep in mind that RMA gives back log2 values. expresso with mas gives back linear scale values. On Thu, 22 May 2003, Andrew Harrison wrote: > Hi, > > I've generated several MA/RI plots after using RMA and > noticed that the standard deviation in M (fold change) is related > to A (average intensity) - to derive the SD in M, only sample fold > changes over a limited range in A (a similar calculation to the > Quackenbush sliding Z-score). > > When using RMA, the standard deviation in M is proportional > to A, except for the largest intensities. However, when using > expresso with all the options set to MAS, the standard deviation is > independent of A. I believe the variance in M should be independent > of A, in which case this suggests MAS is the way to go. > > Has anybody else seen this effect, or can suggest a remedy, > or more likely a gap in my logic? > > Best wishes and thanks in advance, > Harry > >