differential expression analysis-- paired-end and single-end

Question

[Jennifer]

Is it possible to compare gene expressions of two types of sequencing data, i.e. treated samples that are paired-end 50bp and untreated samples that are single-end 40bp? My idea is to get read counts from HTSeq-count, followed by DESeq2. HTSeq-count counts paired-end data as read pairs, therefore an aligned paired-end reads will only be counted as one. Because of this, I think I can do DE with paired end vs single end data?

Thanks! Jenn

Answer 1

[GouthamAtla]

The paired-end reads should be counted only ones as one fragment.

You can include the library type in the design formula in DESeq2/edgeR. Something like design <- model.matrix(~libType+treat) in DESeq2

Otherwise, you can also use RUVSeq or other tools to remove any "batch" effects because of the lib type.