Hi All,

I have three conditions and two replicates per condition. I'd like to make three comparisons:

(1) TE vs [MP and EE] (2) MP vs [TE and EE] (3) EE vs [TE and MP]

I've tried doing the comparison below but output doesn't seem quite right.

Could someone point me how to correctly make these comparisons and where in my code below I might be making a mistake. Any help would be greatly appreciate, many thanks.

Code should be placed in three backticks as shown below

samples <- c('TE_1', 'TE_2',
             'MP_1', 'MP_2',
            'EE_1','EE_2')
replicates <- c('TE', 'TE',
                'MP','MP',
               'EE','EE')

col_data = matrix(replicates, # the data elements 
   nrow=6,              # number of rows 
   ncol=1,              # number of columns 
   byrow = TRUE)

colnames(col_data) <- c('condition')
rownames(col_data) <- samples
col_data <- as.data.frame(col_data)

dds <- DESeqDataSetFromMatrix(countData = counts,
                              colData = col_data,
                              design = ~ condition)

dds <- DESeq(dds)

TE_results <- results(dds, contrast = c(1, -1/2, -1/2))
MP_results <- results(dds, contrast = c(-1/2, 1, -1/2))
EE_results <- results(dds, contrast = c(-1/2, -1/2, 1))

If you have a design of ~0 + condition this will help. That is the type of design you appear to be forming numeric contrasts for, as opposed to the reference-level-based design ~condition.