Question: LIMMA 2x2 with time
0
2.3 years ago by
Ahdee40
United States
Ahdee40 wrote:

Hi can someone take a look at this and see if my design is correct?  What I'm trying to achieve is similar to a 2-way ANOVA main effect and interactions with LIMMA.

I would like to look at the main effect of gene, time and interaction as well as comparison of gene regardless of time. Sorry if this has been asked before, but  I was hoping to get some confirmation.

thanks!

ds <- read.table(header = TRUE, text =                  "id gene time                    1 gfp t24                   2 ko   t24                   3 gfp t75                   4 ko t75                  ")

ft <- paste(ds$time,ds$gene,sep=".")

ft <- factor(ft) ft design2 <- model.matrix(~0+ft) colnames(design2) <- levels(ft) design2

cont.matrix <- makeContrasts(   gene.main = "(t24.ko + t75.ko)/2 - (t24.gfp + t75.gfp)/2",    time.main = "(t75.gfp + t75.ko)  /2  - (  t24.gfp +  t24.ko) /2 ",   Interaction="(t24.ko - t75.ko) - (t24.gfp - t75.gfp)",   gfp.vs.ko = " (t24.ko + t75.ko) - (t24.gfp + t75.gfp) "   ,levels=design2)

limma limma design matrix • 343 views
modified 2.3 years ago by James W. MacDonald51k • written 2.3 years ago by Ahdee40
0
2.3 years ago by
United States
James W. MacDonald51k wrote:

You need to divide by two for the last contrast as well, but otherwise seems ok to me. Plus I don't think you want to use quotes

Thanks James! After reading over you comment I just realized that what I want is indeed the same as the main effect for gene.main.  great thanks again.