Dear list, I've tried to use hist (affy package) to check the quality of two different sets of CEL files from the same experiment: * The original set of CEL files. * The permuted set of CEL files (raw signal intensities for each probe pairs were randomly permuted for each array to create uninformative CEL files) When hist function is called using the raw original data and the raw permuted data separately, the histograms of both set of CEL files are the same, even if the probe-pair values are randomly changed in the second set. This fact suggests that the density graph is built using the probe-pair values instead of probe-set values. But, after a preprocessing method is used on both sets (for example, VSN), the results are quite different: the histograms are not the same and the permutation can be observed easier in the second set. So, this suggests that: * The hist function builds the density graph using the probe pair values when an Affybatch object is used (that is, raw data). * The hist function builds the density graph using the probe set values when an expression set object is used (that is, after preprocessing). Is that correct? Thanks in advance, Javier [[alternative HTML version deleted]]

> * The hist function builds the density graph using the probe pair > values when an Affybatch object is used (that is, raw data). > * The hist function builds the density graph using the probe set > values when an expression set object is used (that is, after > preprocessing). > >Is that correct? Yes. The Affybatch object only has probe-level data in it, so that is what is used for the density graph. You can only get probe set data after applying some algorithm to the probe-level data, and an ExpressionSet object contains only the probe set data. Cheers, Jenny >Thanks in advance, >Javier > > > [[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 Jenny Drnevich, Ph.D. Functional Genomics Bioinformatics Specialist W.M. Keck Center for Comparative and Functional Genomics Roy J. Carver Biotechnology Center University of Illinois, Urbana-Champaign 330 ERML 1201 W. Gregory Dr. Urbana, IL 61801 USA ph: 217-244-7355 fax: 217-265-5066 e-mail: drnevich at illinois.edu

Hi Javier, Javier P?rez Florido wrote: > Dear list, > I've tried to use hist (affy package) to check the quality of two > different sets of CEL files from the same experiment: > > * The original set of CEL files. > * The permuted set of CEL files (raw signal intensities for each > probe pairs were randomly permuted for each array to create > uninformative CEL files) > > When hist function is called using the raw original data and the raw > permuted data separately, the histograms of both set of CEL files are > the same, even if the probe-pair values are randomly changed in the > second set. This fact suggests that the density graph is built using the > probe-pair values instead of probe-set values. > > But, after a preprocessing method is used on both sets (for example, > VSN), the results are quite different: the histograms are not the same > and the permutation can be observed easier in the second set. So, this > suggests that: > > * The hist function builds the density graph using the probe pair > values when an Affybatch object is used (that is, raw data). > * The hist function builds the density graph using the probe set > values when an expression set object is used (that is, after > preprocessing). > > Is that correct? Of course! The AffyBatch object contains only raw data and the ExpressionSet contains only processed data, so what else could be used in either case? Best, Jim > > Thanks in advance, > Javier > > > [[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 -- James W. MacDonald, M.S. Biostatistician Douglas Lab University of Michigan Department of Human Genetics 5912 Buhl 1241 E. Catherine St. Ann Arbor MI 48109-5618 734-615-7826