For now, I impemented it in a way that it picks up where it failed, so the expensive stuff is not repeated, but the problem can possibly be avoided by first getting a prefetch file (see https://twitter.com/lh3lh3/status/1779876367200387172). Also, fasterq-dump could be used to speed things up.
@claraptzsl you may want to look into this, otherwise, you have to restart the pipeline many times on a larger dataset.
The rule file https://github.com/KatharinaHoff/braker-snake/blob/main/rules/rnaseq_download.smk contains a rule download_fastq, which is prone to repeated failures. These are caused by connection timeouts.
For now, I impemented it in a way that it picks up where it failed, so the expensive stuff is not repeated, but the problem can possibly be avoided by first getting a prefetch file (see https://twitter.com/lh3lh3/status/1779876367200387172). Also, fasterq-dump could be used to speed things up.
@claraptzsl you may want to look into this, otherwise, you have to restart the pipeline many times on a larger dataset.