Multifactor design with Genotype: Is it possible to group all?
Treat Time
Sample1 Placebo 0h
Sample2 Placebo 0h
Sample3 Placebo 1h
Sample4 Placebo 1h
Sample7 Drug 0h
Sample8 Drug 0h
Sample9 Drug 1h
Sample10Drug 1h
For an experiment with above factors, I identified DEGs,
Group <- factor(paste(targets$Treat,targets$Time,sep="."))
cbind(targets,Group=Group)
design <- model.matrix(~0+Group)
colnames(design) <- levels(Group)
fit <- glmQLFit(y, design)
my.contrasts <- makeContrasts(
+ Drug.1vs0 = Drug.1h-Drug.0h,
+ Placebo.2vs0 = Placebo.2h-Placebo.0h,
+ levels=design)
qlf <- glmQLFTest(fit, contrast=my.contrasts[,"Drug.1vs0"])
But when new factor is added, for ex, Genotype,
Treat Time Genotype
Sample1 Placebo 0h a
Sample2 Placebo 0h a
Sample3 Placebo 1h a
Sample4 Placebo 1h a
.
.
Sample4 Placebo 1h c
Sample7 Drug 0h c
Sample8 Drug 0h c
Sample9 Drug 1h c
Sample10Drug 1h c
Can I combine the "Genotype" factor and follow the steps as above?
Group <- factor(paste(targets$Treat,targets$Time,targets$Genotype,sep="."))
My aim is to extract genes that are induced at 1hr of drug treatment only for Genotype b (not a and c)
• 142 views
•
link
0 answers
No answers yet.
Log in to answer this question.
I realize the I should have had posted this topic in Bioconductor site. Its my mistake. How can I deal with it without cross-posting?
Mods can comment best, but at minimum you can add a direct link to the cross-post here (and vice versa) so that folks know it's in both fora. Like on this post here: A: Building a biologically meaningful design matrix for limma/voom in a large RNASe.
I mean I didnt post it on Bioc yet. But wanted to see users in that forum to see this post.