On Thu, May 20, 2010 at 11:57 AM, varpal singh <gilvarpal@gmail.com> wrote: > Respected Sir, > Sory sir, am distruing you again, but my all results depend on it. > I'm sending you two outputs that I got, Just, I would like to know which > contrast is rigt, in two condition only ?? > *c<-makeContrasts(group2-group1,levels=design) OR ** > c<-makeContrasts(group1-group2,levels=design) > *Because i get result opposite in both cases, that you can see in these > two probe below *253416_at, 266294_at > FIRST* > library(affy) > data <- ReadAffy() > design<- model.matrix(~0+factor(c(1,1,2,2))) > colnames(design)<-c("group1", "group2") > *c<-makeContrasts(group2-group1,levels=design)* > eset <- rma(data) > fit <- lmFit(eset,design) > fit2 <- contrasts.fit(fit,c) > fit2 <- eBayes(fit2) > topTable(fit2, coef=1, adjust="fdr", number=10) > ID logFC AveExpr t P.Value > adj.P.Val B > 8516 253416_at *7.122005* 7.535481 26.57534 2.255100e-05 0.3000238 > 1.4319777 > 21394 266294_at *5.221736* 7.090044 24.38216 3.100325e-05 0.3000238 > 1.3731773 > > -------------------------------------------------------------------- ----------------------- > * > > SECOND* > design<- model.matrix(~0+factor(c(1,1,2,2))) > colnames(design)<-c("group1", "group2") > *c<-makeContrasts(group1-group2,levels=design)* > fit2 <- contrasts.fit(fit,c) > fit3 <- eBayes(fit2) > topTable(fit3, coef=1, adjust="fdr", number=10) > ID logFC AveExpr t P.Value > adj.P.Val B > 8516 253416_at *-7.122005* 7.535481 -26.57534 2.255100e-05 0.3000238 > 1.4319777 > 21394 266294_at* -5.221736* 7.090044 -24.38216 3.100325e-05 0.3000238 > 1.3731773 > > Hi, Varpal. They should be opposite, yes. Otherwise, the results should be identical and it appears that they are. Sean [[alternative HTML version deleted]]

On Thu, May 20, 2010 at 1:10 PM, varpal singh <gilvarpal@gmail.com> wrote: > Thanks very much for reply, > > I would like to know which contrasts is appropriate for find differential > expression genes ?? > > Because my aim to find significant genes which have fold change value less > than or greater than 2, in both cases have varibility in up regulation and > down regulation gene > > So that, am confuse to select which fold change value, i can select for my > further analysis ? > > Plz tell me which contrast I wil take ?? > > Hi, Varpal. The contrasts are equivalent. Please do use the bioconductor list--you will get the best service there. Sean > Varpall > > > On Thu, May 20, 2010 at 10:23 PM, Sean Davis <seandavi@gmail.com> wrote: > >> >> >> On Thu, May 20, 2010 at 11:57 AM, varpal singh <gilvarpal@gmail.com>wrote: >> >>> Respected Sir, >>> Sory sir, am distruing you again, but my all results depend on it. >>> I'm sending you two outputs that I got, Just, I would like to know which >>> contrast is rigt, in two condition only ?? >>> *c<-makeContrasts(group2-group1,levels=design) OR ** >>> c<-makeContrasts(group1-group2,levels=design) >>> *Because i get result opposite in both cases, that you can see in these >>> two probe below *253416_at, 266294_at >>> FIRST* >>> library(affy) >>> data <- ReadAffy() >>> design<- model.matrix(~0+factor(c(1,1,2,2))) >>> colnames(design)<-c("group1", "group2") >>> *c<-makeContrasts(group2-group1,levels=design)* >>> eset <- rma(data) >>> fit <- lmFit(eset,design) >>> fit2 <- contrasts.fit(fit,c) >>> fit2 <- eBayes(fit2) >>> topTable(fit2, coef=1, adjust="fdr", number=10) >>> ID logFC AveExpr t P.Value >>> adj.P.Val B >>> 8516 253416_at *7.122005* 7.535481 26.57534 2.255100e-05 0.3000238 >>> 1.4319777 >>> 21394 266294_at *5.221736* 7.090044 24.38216 3.100325e-05 0.3000238 >>> 1.3731773 >>> >>> ------------------------------------------------------------------ ------------------------- >>> * >>> >>> SECOND* >>> design<- model.matrix(~0+factor(c(1,1,2,2))) >>> colnames(design)<-c("group1", "group2") >>> *c<-makeContrasts(group1-group2,levels=design)* >>> fit2 <- contrasts.fit(fit,c) >>> fit3 <- eBayes(fit2) >>> topTable(fit3, coef=1, adjust="fdr", number=10) >>> ID logFC AveExpr t P.Value >>> adj.P.Val B >>> 8516 253416_at *-7.122005* 7.535481 -26.57534 2.255100e-05 0.3000238 >>> 1.4319777 >>> 21394 266294_at* -5.221736* 7.090044 -24.38216 3.100325e-05 0.3000238 >>> 1.3731773 >>> >>> >> Hi, Varpal. >> >> They should be opposite, yes. Otherwise, the results should be identical >> and it appears that they are. >> >> Sean >> > > [[alternative HTML version deleted]]