Blocking and using contrasts in voom (limma) (RNAseq)

[Travis Lawrence]

I am having trouble making the contrast matrix when I am using library preparation type (e.g. paired-end, single-end) as an addition variable in the model matrix:

Treatment <- factor(targets$Treatment[c(1:8,11:13)], levels=c("T1", "T2", "T3", "T4"))
Seq <- factor(targets$Sequence[c(1:8,11:13)], level=c("single", "pair"))
design <- model.matrix(~SeqType + Treatment)
y <- voom(x, design, plot=TRUE)
fit <- lmFit(y, design)

Where SeqType is a factor with a value for each RNAseq sample of either paired or single and Treatment is a factor labeling each sample as one of four treatments.

The resulting design matrix for the fit object is:

(Intercept) Seqpair Treatment2 Treatment3 Treatment4
      1       1        0          1          0  
      1       1        0          0          1
      1       0        1          0          0 
      1       1        1          0          0
      1       1        0          0          0
      1       0        0          0          1
      1       0        0          1          0
      1       0        0          0          0
      1       1        1          0          0
      1       1        0          1          0
      1       1        0          0          1

I am unsure of how to construct the contrast matrix to test all pairwise comparisons of treatments.

I have previously used the makeContrasts command, however, I also previously constructed the design matrix by providing it an intercept (design <- model.matrix(~0 + Treatment)) so constructing the contrasts was done by just giving the column headings of the design matrix to the makeContrasts function:

contrast.matrix <- makeContrasts(T1vsT2 = Treatment1-Treatment2)