interaction effect (4x2)
1
0
Entering edit mode
@sebastien-gerega-2229
Last seen 8.1 years ago
Hi, I am having trouble setting up the design for a microarray analysis. It involves 40 samples that are split into 4 groups and are treated in one of 2 ways. What I want to do is identify genes with an interaction effect between group and treatment. What would the best way to go about this? I have attempted the following: interDesign = model.matrix(~factor(sDrug) * factor(sGroup)) interFit = lmFit(lumi.N.P, interDesign) interCont = cbind(c(0,0,0,0,0,1,0,0),c(0,0,0,0,0,0,1,0),c(0,0,0,0,0,0,0,1)) interFit = contrasts.fit(interFit, interCont) interFit = eBayes(interFit) interDTest = decideTests(interFit, method="nestedF", adjust.method="fdr", p.value=0.05) which(abs(interDTest[,1]) == 1 | abs(interDTest[,2]) == 1 | abs(interDTest[,3]) == 1) Is this a suitable way to identify the genes with an interaction effect? So far, from looking at expression profiles, I don't seem to be picking out interesting genes.... Any help would be greatly appreciated. thanks, Sebastien
Microarray GO Microarray GO • 569 views
ADD COMMENT
0
Entering edit mode
@james-w-macdonald-5106
Last seen 2 days ago
United States
Hi Sebastien, Sebastien Gerega wrote: > Hi, > I am having trouble setting up the design for a microarray analysis. > It involves 40 samples that are split into 4 groups and are treated in > one of 2 ways. > What I want to do is identify genes with an interaction effect between > group and treatment. > What would the best way to go about this? I have attempted the following: > > interDesign = model.matrix(~factor(sDrug) * factor(sGroup)) > interFit = lmFit(lumi.N.P, interDesign) > interCont = cbind(c(0,0,0,0,0,1,0,0),c(0,0,0,0,0,0,1,0),c(0,0,0,0,0,0,0,1)) > interFit = contrasts.fit(interFit, interCont) > interFit = eBayes(interFit) > interDTest = decideTests(interFit, method="nestedF", > adjust.method="fdr", p.value=0.05) > which(abs(interDTest[,1]) == 1 | abs(interDTest[,2]) == 1 | > abs(interDTest[,3]) == 1) > > Is this a suitable way to identify the genes with an interaction effect? Well, with 4 groups and 2 treatments I get 6 total interactions. Are the three you are testing here the interesting interactions? > So far, from looking at expression profiles, I don't seem to be picking > out interesting genes.... Interesting defined how? The genes you get aren't a priori genes you want to see? Or you aren't getting any significant genes? Best, Jim > Any help would be greatly appreciated. > thanks, > Sebastien > > _______________________________________________ > 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 Affymetrix and cDNA Microarray Core University of Michigan Cancer Center 1500 E. Medical Center Drive 7410 CCGC Ann Arbor MI 48109 734-647-5623
ADD COMMENT
0
Entering edit mode
Thanks for your reply James, James W. MacDonald wrote: > Well, with 4 groups and 2 treatments I get 6 total interactions. Are > the three you are testing here the interesting interactions? > I guess I am interested in all 6 interactions. How would I go about looking at them all? >> So far, from looking at expression profiles, I don't seem to be >> picking out interesting genes.... > > Interesting defined how? The genes you get aren't a priori genes you > want to see? Or you aren't getting any significant genes? > The reason I said that was initially I accidentally performed the analysis without applying the contrast: interDesign = model.matrix(~factor(sDrug) * factor(sGroup)) interFit = lmFit(lumi.N.P, interDesign) interFit = eBayes(interFit) interDTest = decideTests(interFit, method="nestedF", adjust.method="fdr", p.value=0.05) which(abs(interDTest[,6]) == 1 | abs(interDTest[,7]) == 1 | abs(interDTest[,8]) == 1) And the genes I identified that way were interesting to me, based on a quick glance at expression profiles. Then I realised I should have applied a contrast. thanks again, Sebastien
ADD REPLY
0
Entering edit mode
Hi Sebastien, Sebastien Gerega wrote: > Thanks for your reply James, > > James W. MacDonald wrote: >> Well, with 4 groups and 2 treatments I get 6 total interactions. Are >> the three you are testing here the interesting interactions? >> > > I guess I am interested in all 6 interactions. How would I go about > looking at them all? sD <- factor(sDrug) sG <- factor(sGroup) design <- model.matrix(~0 + sD:sG) Then make a contrasts matrix. Best, Jim > >>> So far, from looking at expression profiles, I don't seem to be >>> picking out interesting genes.... >> >> Interesting defined how? The genes you get aren't a priori genes you >> want to see? Or you aren't getting any significant genes? >> > The reason I said that was initially I accidentally performed the > analysis without applying the contrast: > > interDesign = model.matrix(~factor(sDrug) * factor(sGroup)) > interFit = lmFit(lumi.N.P, interDesign) > interFit = eBayes(interFit) > interDTest = decideTests(interFit, method="nestedF", > adjust.method="fdr", p.value=0.05) > which(abs(interDTest[,6]) == 1 | abs(interDTest[,7]) == 1 | > abs(interDTest[,8]) == 1) > > And the genes I identified that way were interesting to me, based on a > quick glance at expression profiles. Then I realised I should have > applied a contrast. > thanks again, > Sebastien -- James W. MacDonald, M.S. Biostatistician Affymetrix and cDNA Microarray Core University of Michigan Cancer Center 1500 E. Medical Center Drive 7410 CCGC Ann Arbor MI 48109 734-647-5623
ADD REPLY

Login before adding your answer.

Traffic: 391 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6