I'm just asking here whether doing batch correction in a design with 3 batches where one of the batches only contains one sample is possible. From this post Batch Correction with only one sample in a batch#:~:text=The%20bottom%20line%20is%20that,that%20sample%20is%20completely%20uninformative it seems like that sample would just be removed from the analysis, however there isn't any messages of that while running Deseq so I'm just asking for clarity whether the sample is removed or kept in the analysis.
This is the experiment design
BATCH SEX RESPONSE
1 F A
1 F A
2 F A
2 F A
2 F A
2 M A
1 F B
1 M B
3 M B
2 F B
2 M B
2 M B
2 F C
2 M C
2 M C
2 M C
2 M C
And this is the deseq design
dds <- DESeqDataSetFromMatrix(countData = count_matrix,
colData = coldata,
design = formula(paste("~", " batch", "+", " sex", "+", comparison_variable)))
dds <- DESeq(dds)
DE_output <- lfcShrink(dds, coef=paste0(Condition,"_",A_vs_B), type="apeglm")
Yeah, I thought that'd be the case.
Thanks for the suggestion, I'll look into using limma-voom.