[BioC] enquiry about the limma script

Gordon K Smyth smyth at wehi.EDU.AU
Mon Mar 25 23:00:50 CET 2013


Dear Yong Li,

The formatted table you have asked for is not one produced automatically 
by limma.  The closest is that produced by write.fit().  If you are trying 
to reproduce a previous analysis, you should contact the person who 
produced that the previous analysis.

Best wishes
Gordon

> Date: Sun, 24 Mar 2013 20:25:20 +0800
> From: yong li <yonglinet at gmail.com>
> To: <bioconductor at stat.math.ethz.ch>
> Subject: [BioC] enquiry about the limma script
>
> Dear Sir/Madam,
>
>
>
> I meet some problems when I use the limma package to analysis the Agilent
> single channel microarray. I got 8 samples (with 8*60K array) results as
> following:
>
> Samples: (T1, T2, T3) =Treat1
>
> Samples: (T4, T5, T6)=Treat2
> Samples: (C1, C2)=Control
> my targets.txt contain:
> file_name conditon
> x1.txt T1
> x2.txt T2
> x3.txt T3
> x4.txt T4
> x5.txt T5
> x6.txt T6
> x7.txt C1
> x8.txt C8
>
>
> I want to get the different expressed genes under 3 conditions: Treat1 vs
> Control, Treat2 vs Control, Treat2 vs Treat1 with the following format:
>
> *GROUP1*
>
> *GROUP2*
>
> *CONTROL*
>
> *GROUP1-vs-CONTROL*
>
> *GROUP2-vs-CONTROL*
>
> *GROUP1-vs-GROUP2*
>
> *Probeid*
>
> *Accession*
>
> *T1*
>
> *T2*
>
> *T3*
>
> *T4*
>
> *T5*
>
> *T6*
>
> *C1*
>
> *C2*
>
> *avg-group1*
>
> *avg-group2*
>
> *avg-control*
>
> *FC*
>
> *UP/DOWN*
>
> *P value*
>
> *FC*
>
> *UP/DOWN*
>
> *P value*
>
> *FC*
>
> *UP/DOWN*
>
> *P value*
>
> CUST_PI5
>
> AB91
>
> 3
>
> 1
>
> 1
>
> 1
>
> 2
>
> 4
>
> 6
>
> 6
>
> 2
>
> 2.5
>
> 6
>
> 6
>
> DOWN
>
> 0.01
>
> 2
>
> DOWN
>
> 0.02
>
> 1
>
> DOWN
>
> 0.2
>
>
>
> My R script as follow:
>
> library(limma)
>
> targets <- readTargets("targets.txt")
>
> x <- read.maimages(targets[,"file_name"], source="agilent",green.only=TRUE)
>
> y <- backgroundCorrect(x, method="normexp")
>
> y <- normalizeBetweenArrays(y, method="quantile")
>
> y.ave <- avereps(y, ID=y$genes$ProbeName)
>
> f <- factor(targets$condition, levels = unique(targets$condition))
>
> design <- model.matrix(~0 + f)
>
> colnames(design) <- levels(f)
>
>
>
> fit <- lmFit(y.ave, design)
>
> contrast.matrix <- makeContrasts("Treat1-C",
> "Treat2-C","Treat2-Treat1",levels=design)
>
> fit2 <- contrasts.fit(fit, contrast.matrix)
>
> fit2 <- eBayes(fit2)
>
> write.table(topTable(fit2,adjust="BH",number="44000"),
> file="out_file.txt",sep="\t")
>
>
>
> I got the results as following:
>
> *ProbeName*
>
> *SystematicName*
>
> *Treat1.C*
>
> *Treat2.C*
>
> *Treat1.Treat2*
>
> *AveExpr*
>
> *F*
>
> *P.Value*
>
> *adj.P.Val*
>
> CUST_248
>
> EV478
>
> -0.54306
>
> -8.90559
>
> 8.362525
>
> 14.51936
>
> 2562.081
>
> 1.97E-10
>
> 6.12E-06
>
>
> The problem is that I can’t get the normalized data, every group average
> data, FC with up/down label in a file just like the previous format. I
> write you for help and look forward to receiving your reply.
>
> Thanks a lot.
>
>
>
> Yong Li
>
> Institute of Plant Physiology and Ecology, SIBS, CAS
>

______________________________________________________________________
The information in this email is confidential and intend...{{dropped:5}}


More information about the Bioconductor mailing list