trimming 151bp to shorter k-mer in ATAC-seq analysis

Question

[QX]

Hi all,

I have dataset ATAC-seq for 151bp and process to the analysis. However, I read in some tutorial that "fragment length distribution of ATAC-seq libraries usually peaks at around 45–50 bp , and as a result, a majority of fragments are already fully covered by 2x36mer reads." For this reason, shall I cut down the read from 151 to a shorter length (like 65 or 35) before proceed the mapping?

[rfran010]

Yes, read trimming is a normal step in ATAC-seq workflow, although I don't know that a majority of fragments are covered by 36mer reads.

You can try cutadapt, fastp, or another tool to trim reads with read-through.

[QX]

yes I do trimming, but usually for adapters only. Do we need to trim to a specific number, for e.g. 36?

[ATpoint]

No, trim for adapters and call it a day.

[QX]

so only trim for adapters and the length remain does not matter? do you know why they mention the term '2x36mer reads', does it make any biological sense of ATAC-seq?