Hi

We have one batch which was done with single ended illumina reads and the rest paired. I ran salmon paired end mode for the other batches then changed to single ended mode for the singled ended reads. I then imported all the data together using tximport. I then did VST using DESEQ2 and did a PCA and the single ended reads are in their own cluster away from the rest of the samples. So it seems there is a batch effect, is this to be expected?

Kind Regards,

Chris

Yes, it is always to be expected. Any two different sequencing technologies (different read lengths, single vs paired, different amplification strategies, or whatever) will yield a batch effect.