Different trimming for different batches within the same experiment

Question

[Mozart]

Hello there, I have to compare datasets gathered from different labs. one lab sequenced their samples in singled-end mode while other two labs did a paired-end sequencing. Looking at the fastq files, batch from one lab contains minimal adapter contamination whereas the batches from the other labs contains no contamination at all. not sure if it's best to trim all the files, regardless? this, in order to have a comparable level of clipping?

Answer 1

[ATpoint]

From all the biases you have when comparing independent studies the adapter trimming is probably the least relevant. Just trim adapters if necessary, and then go ahead with downstream. Other batches such as library prep/strandedness of the kit, RNA extraction, sequencing depth and sample size will be the factors driving batch effects.