Dear Dr.Smyth and Limma users; Good morning. I have an experiment which has 4 variables; sham surgery, full surgery, treatment 1 and treatment 2. I am having difficulty formulating a design matrix and interaction terms for this experiment. Treatment 1 is known to us and treatment 2 is the novel therapeutic. Groups are as follows: sham + treatment1 sham + treatment2 full + treatment1 full + treatment2 I need to answer: * gene exp difference between treat1 and treat2 in both models sham and full in order to verify full surgery working, and to prove treatment 2 is working to protect animal from full surgery detriments. Treatment1 is considered a known drug with known effects. I am also confused (or ignorant) on how to employ a proper "difference of difference" or interaction term. I greatly appreciate any help and understand and respect everyone's time constraints. Thank you, Brian -- output of sessionInfo(): Q no code -- Sent via the guest posting facility at bioconductor.org.

Hi Brian, A completely analogous example is in the limma User's Guide, section 8.5.2. Let us know if that doesn't answer your question. Best, Jim On Friday, September 06, 2013 10:40:43 AM, Brian [guest] wrote: > > Dear Dr.Smyth and Limma users; > > Good morning. > I have an experiment which has 4 variables; sham surgery, full surgery, treatment 1 and treatment 2. I am having difficulty formulating a design matrix and interaction terms for this experiment. Treatment 1 is known to us and treatment 2 is the novel therapeutic. > > Groups are as follows: > sham + treatment1 > sham + treatment2 > full + treatment1 > full + treatment2 > > I need to answer: > * gene exp difference between treat1 and treat2 in both models sham and full in order to verify full surgery working, and to prove treatment 2 is working to protect animal from full surgery detriments. Treatment1 is considered a known drug with known effects. > > I am also confused (or ignorant) on how to employ a proper "difference of difference" or interaction term. > > I greatly appreciate any help and understand and respect everyone's time constraints. > Thank you, > Brian > > -- output of sessionInfo(): > > Q no code > > -- > Sent via the guest posting facility at bioconductor.org. > > _______________________________________________ > 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 -- James W. MacDonald, M.S. Biostatistician University of Washington Environmental and Occupational Health Sciences 4225 Roosevelt Way NE, # 100 Seattle WA 98105-6099