Hi all, I'm still in trouble with reading data from a text file. I worked on CEL files and run the rma function on them. Then it saved the results as a test file. Now, I want to look at the plots (scatter and boxplot), but I don't want to do them one by one. I know that there must be an easy way to get all the plots. Could anyone help me with that. Then, I wanted to look at the differentially expressed genes using th elinear model discribed in chapter 7.2 of the limma notes on linear models, but I don't understand how the model.matrix function works and what "~ -1+factor(c(...)))" means. More precisely, I don't know how I can use it with my data. I have 3 different microarray data that I want to compare. Each microarray is a time serie of 6 time points. I don't know how I should write my vector. The way I wrote it seemed ok, but it gave me NA all the way. Why? I thought BioConductor was supposed to be easy to use, but I'm getting lost because I don't know what I can do with it and then how to do it. Is there a documentation covering the ploting part and how to look for differentially expressed genes? I know it's many questions, but I really need to understand and learn how to use this very usefull tool. Thanks a lot in advance, Jamila

On Mon, Mar 08, 2004 at 03:31:40AM -0800, Jamila Ahdidan wrote: > Hi all, > I'm still in trouble with reading data from a text > file. I worked on CEL files and run the rma function > on them. Then it saved the results as a test file. > Now, I want to look at the plots (scatter and > boxplot), but I don't want to do them one by one. I > know that there must be an easy way to get all the > plots. Could anyone help me with that. Yes, there are commands in R for boxplots and scatter plots. Have you looked at any of the vignettes for affy? Start R > library(affy) > openVignette() then select the one that says affy primer and on page 15 it seems that there are some descriptions. > > Then, I wanted to look at the differentially expressed > genes using th elinear model discribed in chapter 7.2 > of the limma notes on linear models, but I don't > understand how the model.matrix function works and > what "~ -1+factor(c(...)))" means. More precisely, I > don't know how I can use it with my data. I have 3 > different microarray data that I want to compare. Each > microarray is a time serie of 6 time points. I don't > know how I should write my vector. The way I wrote it > seemed ok, but it gave me NA all the way. Why? > > I thought BioConductor was supposed to be easy to use, > but I'm getting lost because I don't know what I can > do with it and then how to do it. Is there a > documentation covering the ploting part and how to > look for differentially expressed genes? Well, that depends a lot on what you already know. Bioconductor is based on R, those that are familiar with R and statistics will find it easier than those that are not. But your best strategy is to find the vignettes and to read them carefully. There are some other tools (in the tkWidgets package) called vExplorer that will help you to step through the commands in a vignette. As for fitting linear models, there are very many alternatives. Limma is one but genefilter may be more appropriate for what you are doing. Also the factDesign library is set up to handle factorial experiments. For time series experiments you may want to think more carefully about what an appropriate single gene analysis might be and then extrapolate that to the microarray setting. Robert > > I know it's many questions, but I really need to > understand and learn how to use this very usefull > tool. > > Thanks a lot in advance, > Jamila > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor -- +--------------------------------------------------------------------- ------+ | Robert Gentleman phone : (617) 632-5250 | | Associate Professor fax: (617) 632-2444 | | Department of Biostatistics office: M1B20 | | Harvard School of Public Health email: rgentlem@jimmy.harvard.edu | +--------------------------------------------------------------------- ------+

Jamilia, If you did this - data <- ReadAffy() eset <- rma(data) write.exprs(eset,file="myresults.txt") Then you can get the data back in by - rma <- read.table("C:/your directory/myresults.txt", header=T) You can then get a boxplot by boxplot(rma) boxplot(rma[,1]) #to select a specific column a scatterplot plot(rma[,1], rma[,2], pch=".") If you mean you want boxplots of the CEL files before rma then data <- ReadAffy() boxplot(data) If you have an exprsset eset <- rma(data) Then to get a boxplot of this you could try (example with 6 chips) x <- exprs(eset) boxplot(x[,1], x[,2], x[,3], x[,4], x[,5], x[,6]) You might find also the following useful as you can add phenodata (exp, time, replicate etc..) as covariates which you can then subset on later eg: eset$experiment data <- ReadAffy(widget=T) I'm also new to BioC/R but find the html help for the packages as well as vignettes useful. Good luck Matt