This is simple! Thanks!

This is super easy, but you still have to extract the FASTQ files from the SRA file using fastq-dump. So it might be just easier to just use fastq-dump, changing the default save directory, according to piet's post.

[UPDATE] Funny enough, just because I posted this, I changed the configuration as by piet's post (which it has worked for me in the past), but now files are still being saved to my home directory for no reason.. gah... downloading SRA files and extracting fastq locally using fastq-dump will probably be the way to move forward.

Unfortunately, NCBI recommends against directly downloading files as of 2019. Basically, prefetch everything.

Quick warning in case memory is a potential issue: I believe fastq-dump downloads an sra file to the deafult sra directory, converts it to fastq, then deletes the sra file.

I executed below, where fetch is a file with each line consisting of an accession number.

while read -r line
do
    accession_number="$line"
    fastq-dump --split-files --gzip $accession_number
done < ${fetch}

A couple of my jobs failed due to cluster memory issues, so I checked my ~/ncbi/public/sra directory, and, lo and behold, tons of sra files.

So complicated! Thanks!