pseudobulk DESeq2 design matrix

Question

[HK]

Hi All,

I have the following situation and I just want to make sure that I understand everything correctly from a statistics point of view.

I have Tissue 1 and Tissue 2 (same patient), so no batch effect as same patient and sequencing was done on the same time.

patient condition
p1  Tisuue1
p2  Tissue 1
p3  Tisuue1
p1  Tissue 2
p2  Tisuue2
p3  Tissue 2

dds <- DESeqDataSetFromMatrix(countData = round(counts_cellType),
                              colData = colData,
                              design = ~ condition)
# filter
keep <- rowSums(counts(dds)) >=10
dds <- dds[keep,]
# run DESeq2
dds <- DESeq(dds)

Now I have tissue 3 and tissue 4 (public datasets), so different time, different number of cells etc.

I want to compare Tissue1 to Tissue 3 and Tissue4.

patient condition
p1  Tisuue1
p2  Tissue 1
p3  Tisuue1
A1  Tissue 3
A2  Tisuue3
A3  Tissue 3
B1  Tissue 4
B2  Tissue 4
B3  Tissue 4

How should I cater this analysis using DEseq2? These are single cell datasets, and I am performing Differential analysis per cell type.