Process each sample in its own folder
The previous release was inadvertently deleting the reference database after the first sample. This release will process each sample in its own folder, and only delete the reference database once the entire process is complete.
Also clear the fastq-dump cache after each sample