Hi,

I have 2545 and 1402 genes in Oncology Biomarker Panel and Precision Immuno-oncology Panel respectively. I have 719 common genes between two panels. I will need to merge raw read counts from these panels for differential expression (we know experimental condition, chemistry for samples from the same patients makes this reasonable). However, I thought to do differential expression analysis for 719 common by taking panel effect as batch like below

    condition   batch
A1  treatment   1
A2  treatment   1
A3  treatment   1
A4  treatment   1
A5  treatment   1
A6  treatment   1
A7  control     1
A8  control     1
A9  control     1
A10 control     1
A11 control     1
A12 control     1
B1  treatment   2
B2  treatment   2
B3  treatment   2
B4  treatment   2
B5  treatment   2
B6  treatment   2
B7  control     2
B8  control     2
B9  control     2
B10 control     2
B11 control     2
B12 control     2 

dds <- DESeqDataSetFromMatrix(countData = countData,colData = colData,design = ~ batch + condition)

neu.dds.LRT <- DESeq(dds,betaPrior=FALSE, test="LRT",
                     full=~ batch + condition, reduced=~batch)
dpsc.res.LRT <- results(neu.dds.LRT , contrast=c("condition", "treatment", "control"))
)

So that I would have differentially expressed genes among 719 common genes for treatment Vs control considering panel effect. Then I should combine uncommon genes between panels and do differential expression analysis separately (no longer with batch correction). I would have two sets of results. My question is , is there any way to have a set of differential expression analysis with two panels raw read counts although they have 719 common enes? Or, is these any way to combine design after making dds to unified my results?

Thanks for any assistance

What technology are you using here?

I'm not sure at all about the idea of merging counts from different platforms, and whether taking the intersection and allowing for a shift is sufficient to deal with the systematic differences.