I personally would not remove control transcripts before looking for differential expression, since they may serve as a negative control, i.e. if many of them are differentially expressed between samples it means that something is wrong (or at least suspicious). Best regards, Moshe. > Dear list, > A possible data analysis workflow for EXON arrays could be as follows > (extracted from "Exon Array data analysis using Affymetrix Power Tools > and R statistical software", Briefings in Bioinformatics): > > * Normalization and summarization (at exon or gene-level) of the > array set. > * Quality control of exon array data of summarization results (to > remove possible outliers) > * Specific filtering steps, for example: > o Restrict analysis to core probesets > o Filter for undetected probesets (i.e., undetected exons), > making use of DABG (Detected above background) analysis. > o Filter for cross-hybridizing probesets (exons) > o Filter for genes undetected genes in all groups > > I'm running a gene-level data analysis on Human GENE ST 1.0 (not EXON) > arrays, which are, in principle, designed for gene expression profiling, > that is, a gene-level analysis. My question is related to the filtering > step. I was wondering if, once the normalization and summarization is > run at the transcript level (core), giving 33297 transcripts, the > following filtering can be run before differential expression analysis: > > * Remove control transcripts such as other_spike, AFFX, pos_control > (normgene->exon) and neg_control (normgene->intron). This step > removes around 4156 transcripts > * Remove transcripts with very low variability through varFilter > function (genefilter package) > > Since these were the steps recommended in "Bioconductor case studies" > book for 3'IVT arrays (the controls were different in 3'IVT), I was > wondering if these 2 filtering steps can also be used on Human Gene > arrays for gene-level analysis or, on the contrary, I have to run the > filtering steps described above for EXON arrays. > Thanks, > Javier > P.S. If you know any data analysis workflow document for HuGene arrays, > please, let me know > > [[alternative HTML version deleted]] > > _______________________________________________ > Bioconductor mailing list > Bioconductor at r-project.org > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: > http://news.gmane.org/gmane.science.biology.informatics.conductor > -- Moshe Olshansky Division of Bioinformatics The Walter & Eliza Hall Institute of Medical Research 1G Royal Parade, Parkville, Vic 3052 e-mail: olshansky at wehi.edu.au tel: (03) 9345 2697 ______________________________________________________________________ The information in this email is confidential and intend...{{dropped:4}}