Hello all, My question is rather similar to one that was asked back in 2010 ( http://article.gmane.org/gmane.comp.lang.r.sequencing/1663/match=edgeR +time+serie ) I have an RNA Seq experiment with 2 replicates of 8 time points with both a control and treatment conditions. Time point 0 however represents the culture prior to any experimental treatment (control or test) ---------------------------------------------- Control 4hr,8hr,12hr,16hr,20hr,24hr,48hr 0hr< Treatment 4hr,8hr,12hr,16hr,20hr,24hr,48hr ----------------------------------------------- My previous approach in my analysis was to duplicate the 0hr data point to create a control and treatment variation: ------------------------------------------------- Control 0hr,4hr,8hr,12hr,16hr,20hr,24hr,48hr vs Treatment 0hr,4hr,8hr,12hr,16hr,20hr,24hr,48hr ------------------------------------------------- While this allowed me to get preliminary results it is clearly not the best approach to take. What design and what glm coefficients do I use to properly determine the difference between Control and treatment conditions over time? Below I am placing my previous R code. Please excuse me if you find it to be crude, I am very much a beginner: >counttable <- read.csv("~/Desktop/Countsdata.csv", header=T, row.names=1) >meta <- data.frame(row.names=colnames(counttable),treat=c("A","A","A","A","A", "A","A","A","A","A","A","A","A","A","A","A","B","B","B","B","B","B","B ","B","B","B","B","B","B","B","B","B"),time=c("0","0","1","1","2","2", "3","3","4","4","5","5","6","6","7","7","0","0","1","1","2","2","3","3 ","4","4","5","5","6","6","7","7")) >group <- factor(paste(meta$treat,meta$time,sep=".")) >meta$treat <-relevel(meta$treat,ref="A") >cbind(meta,group=group) >design <- model.matrix(~treat * time, data=meta) >y <- DGEList(counts=counttable, group=group) >y <- calcNormFactors(y) >y <- estimateGLMCommonDisp(y, design) >y <- estimateGLMTrendedDisp(y, design) >y <- estimateGLMTagwiseDisp(y, design) >fit <- glmFit(y,design) >lrt <-glmLRT(fit, coef=10:16) ---------------------------------------------------------------------- ---------- Here is y$samples: group lib.size norm.factors A.0.r1 A.0 6081748 1.0384312 A.0.r2 A.0 17974722 0.9427160 A.1.r1 A.1 8370706 1.0130071 A.1.r2 A.1 6212810 0.9273912 A.2.r1 A.2 10982891 1.1070712 A.2.r2 A.2 12331217 0.9592652 A.3.r1 A.3 10477416 1.1163213 A.3.r2 A.3 10011161 0.9715542 A.4.r1 A.4 7501092 1.0879016 A.4.r2 A.4 11169341 0.9939979 A.5.r1 A.5 4847337 1.1255744 A.5.r2 A.5 8984091 0.9484228 A.6.r1 A.6 4335731 1.1164633 A.6.r2 A.6 11960287 0.9137765 A.7.r1 A.7 38520089 1.0384443 A.7.r2 A.7 17030636 0.7711532 B.0.r1 B.0 6081748 1.0384312 B.0.r2 B.0 17974722 0.9427160 B.1.r1 B.1 6544497 1.0382895 B.1.r2 B.1 19784146 0.9356031 B.2.r1 B.2 10151511 1.0821488 B.2.r2 B.2 11254387 1.0013111 B.3.r1 B.3 11721983 1.1303817 B.3.r2 B.3 5900819 0.9468124 B.4.r1 B.4 8292623 1.0678621 B.4.r2 B.4 15323563 0.9716106 B.5.r1 B.5 2840820 1.0969923 B.5.r2 B.5 4515291 0.9427365 B.6.r1 B.6 11301780 1.0919644 B.6.r2 B.6 15830412 0.9406934 B.7.r1 B.7 33395846 1.0560019 B.7.r2 B.7 19404854 0.7798291 Thank you all in advance for your help, Marcelo Pomeranz Ph.D. Post Doc The Ohio State University [[alternative HTML version deleted]]