Dear mail list members, I'm working with Agilent single color arrays (4 x 44K) wich data were extracted by Feature Extraction software; this software provides some flags measurements in the file. In limma manual the authors showed an example with Agilent single color arrays: Time Course E ects of Corn Oil on Rat Thymus with Agilent 4x44K Arrays - but in this they filtered out control probes and low expressed probes. I couldn't find in the manual if limma takes the flag measurement parameters to process data. If not, does anyone have an example of a function do to that? Best regards, Amanda Freire de Assis, MSc Molecular Immunogenetics Group University of São Paulo Av. Bandeirantes, 3900 14049-900 - Ribeirão Preto - Brazil +55 16 36023246 [[alternative HTML version deleted]]

hello Amanda, That is a script of a function ,( sorry but not comment in this function) . Maybe you find something interesting. let me know produce.differential.analysis <- function(RGobj,   ANNOTATION){     print(paste("Starting the  UNIVARIATE differential analysis for Agilent platform"))     design =  read.csv2("DATA/design.txt", sep="\t", header = TRUE)       for (d in 2:ncol(design)){         #gestion Folder         name_folder = colnames(design)[d]#nom de l'analyse differentielle         print("Process for the design ")         print(name_folder)         print("filtering of data")         condition1 =as.vector(na.omit((design[design[d] == 0,1])))         condition2 =as.vector(na.omit((design[design[d] == 1,1])))         label_condition1 =unique(condition1)         label_condition2 =unique(condition2)         current.sumFlag.label_condition1 = apply(RGobj$Rb[,RGobj$target$Condition == label_condition1], 1, sum)         current.sumFlag.label_condition2 = apply(RGobj$Rb[,RGobj$target$Condition == label_condition2], 1, sum)        #retain only the probes  with flag significatif in more 50 pc of samples        mask_row =   current.sumFlag.label_condition1  > length(as.vector(na.omit((design[design[d] == 0,1])))) /2 & current.sumFlag.label_condition2  > length(as.vector(na.omit((design[design[d] == 0,2])))) /2        mask_col =   c(grep(label_condition1,RGobj$targets$Condition), grep(label_condition2,RGobj$targets$Condition))       #summarization       E <- new("MAList", list(targets=RGobj$targets[mask_col,], genes=RGobj$genes[mask_row,], source=RGobj$source, M=RGobj$Gb[mask_row,mask_col], A=RGobj$G[mask_row,mask_col]))       E.avg <- avereps(E, ID=E$genes$ProbeName)        #creation folder analysis        if (!file.exists(paste("RESULT/",name_folder, sep = ""))){ dir.create(paste("RESULT/",name_folder, sep = ""), showWarnings = TRUE, recursive = FALSE, mode = "0777")}          write.table(E$target, paste("RESULT/",name_folder,"/mappage_Condition_NomFichier.csv", sep = "") , quote = TRUE, sep = ";",  eol = "\n", na = "NA", dec = ".", row.names = FALSE, col.names = TRUE)          design_limma= cbind(est_condition1 =rep(1, length(condition1) + length(condition2)) ,                              est_condition2 =c(rep(0, length(condition1)), rep(1, length(condition2))) )          fit = lmFit(E.avg$A, design_limma)          fit2= eBayes(fit)          rs_table = topTable(fit2, coef= 2, adjust= "BH", number =nrow(fit2$genes))          matrice = cbind(ID = rownames(E.avg), E.avg$A)          rs_table_exp = merge(rs_table, matrice, by.x = "ID", by.y = "ID")          rs_table_exp_annotated = merge(rs_table_exp, ANNOTATION, by.x = "ID", by.y = "ProbeID")          write.table(rs_table_exp_annotated, paste("RESULT/",name_folder,"/rs_differential_analysis.csv", sep = "") , quote = TRUE, sep = ";",  eol = "\n", na = "NA", dec = ".", row.names = FALSE, col.names = TRUE)          MAAndVolcanoPlot(rs_table_exp_annotated, paste("RESULT/",name_folder, sep = "") )     } } Le Jeudi 7 novembre 2013 18h58, "amandafassis@usp.br" <amandafassis@usp.br> a écrit : Dear mail list members, I'm working with Agilent single color arrays (4 x 44K) wich data were extracted by Feature Extraction software; this software provides some flags measurements in the file. In limma manual the authors showed an example with Agilent single color arrays: Time Course E ects of Corn Oil on Rat Thymus with Agilent 4x44K Arrays - but in this they filtered out control probes and low expressed probes. I couldn't find in the manual if limma takes the flag measurement parameters to process data. If not, does anyone have an example of a function do to that? Best regards, Amanda Freire de Assis, MSc Molecular Immunogenetics Group University of São Paulo Av. Bandeirantes, 3900 14049-900 - Ribeirão Preto - Brazil +55 16 36023246     [[alternative HTML version deleted]] _______________________________________________ Bioconductor mailing list Bioconductor@r-project.org https://stat.ethz.ch/mailman/listinfo/bioconductor Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor [[alternative HTML version deleted]]

Dear Amanda, The short answer is that limma takes all the parameters that it needs from the Feature Extraction output. I can't give you any more specific answer, because there is no output produced by the Agilent Feature Extraction software called a "flag measurement parameter". limma allows you to access and use any column of information produced by Feature Extraction, but you need to know the name of the parameter that you want to use. Generally you should not need to access anything other than what limma reads automatically. Agilent arrays are usually good quality, and the limma procedures are robust, so agonising over outliers (if that is what is worrying you) is not usually necessary. Best wishes Gordon > Date: Thu, 7 Nov 2013 16:54:20 -0200 (BRST) > From: amandafassis at usp.br > To: bioconductor at r-project.org > Subject: [BioC] Limma and pre processing Agilent single color arrays > > Dear mail list members, > > I'm working with Agilent single color arrays (4 x 44K) wich data were > extracted by Feature Extraction software; this software provides some > flags measurements in the file. > > In limma manual the authors showed an example with Agilent single color > arrays: Time Course E ects of Corn Oil on Rat Thymus with Agilent 4x44K > Arrays - but in this they filtered out control probes and low expressed > probes. > > I couldn't find in the manual if limma takes the flag measurement > parameters to process data. If not, does anyone have an example of a > function do to that? > > > Best regards, > > > Amanda Freire de Assis, MSc > Molecular Immunogenetics Group > University of S??o Paulo > Av. Bandeirantes, 3900 > 14049-900 - Ribeir??o Preto - Brazil > +55 16 36023246 ______________________________________________________________________ The information in this email is confidential and intend...{{dropped:5}}