Hello, I am doing a RNA-Seq analysis with R using the limma package. I have RNA-Seq data of different mutant lines of a model organism which have two different phenotypes in comparison with the siblings as control. The samples are like that:
- line 1, phenotype 1, mutant and sibling
- line 2, phenotype 1, mutant and sibling
- line 3, phenotype 2, mutant and sibling
- line 4, phenotype 2, mutant and sibling
- line 5, phenotype 2, mutant and sibling
That means line 1 and 2 have the similar phenotype 1 and line 3, 4, and 5 have the similar phenotype 2 which differ from phenotype 1. Let's say phenotype 1 has more cells than the control and phenotype 2 has less cells than the control.
I did the RNA-Seq analysis for each line, but now I want to compare the lines of one phenotype. I want to find the genes which are different in the lines with the same phenotype in comparison to the siblings. My makeContrast command looks like that:
cont.matrix <- makeContrasts(pheno1 = (line1.mut+line2.mut)-(line1.sib+line2.sib), pheno2 = (line3.mut+line4.mut+line5.mut)-(line3.sib+line4.sib+line5.sib), levels=design)
Finally I want to find genes (eg. genes involved in the cell cycle) which are in phenotype 1 up and in phenotype 2 down regulated or the other way around.
cont.matrix <- makeContrasts(pheno1vspheno2 = ((line1.mut+line2.mut)-(line1.sib+line2.sib)) - ((line3.mut+line4.mut+line5.mut)-(line3.sib+line4.sib+line5.sib)), levels=design)
I am not sure if a can do a contrast matrix like this. Do you think this will give me the genes I am interested in?
Thank you very much for your explanation. This helped me a lot!