Difference between clipping and without clipping in Fastq-dump

Question

[CTLong]

Hi all,

I have a simple question about the purpose of the --clip option in fastq-dump. I see that adding it is recommended, but the descriptions about it's function is very vague. Based on my understanding of read clipping, I believe it helps trim off adapter sequences or low quality bases? And if that is the case, does not adding it significantly affect my downstream analysis if I am going to carry out adapter and quality trimming with BBMap anyways? Thanks.

Answer 1

[ATpoint]

Tbh I was not even aware of this option, not sure what the sparse documentation means and never used it. Don't set it. Just use default settings and use the fastq files as-is. Trimming etc should be done by established and well-documented tools. Even better, enter the accessions at sra-explorer.info and get download links directly, circumventing the terrible fastq-dump.

[CTLong]

This is an interesting point. Based on my previous experience, downloading fastq files from sra-explorer takes really long with wget (our HPC cluster do not have aspera installed). Do you have any other suggestions in how to optimise the speed? Also, why is fastq-dump inferior to sra-explorer and does this cause significant impact on the analysis?

[ATpoint]

I would have a hard time believing that fastq-dump would be any faster then downloading fastq files directly from ncbi via the wget links. Anyway, aspera is trivial to install without root access: Setting up Aspera Connect (ascp) on Linux and macOS