Running

Overview

This section goes over how to actually execute the workflow

---

Dry Run

It is highly recommended to start a dry-run first. A dry-run does several things:

1. Allows you to see what steps in the workflow need to be completed
2. Ensures preliminary configuration is set up correctly

A dry-run does not truly execute any component of the pipeline. No results will be generated.

Do the following to perform a dry-run

snakemake --profile slurm --dry-run
(NOTE 1: If you did not namek your CookieCutter profile slurm, replace the above line with its appropriate name)
(NOTE 2: If you receive an error when running this, replace slurm with ~/.config/snakemake/slurm)

After several seconds, many lines should move through the terminal.
It should end with This was a dry-run (flag -n). The order of jobs does not reflect the order of execution.

If this is not the case, an error has occured, and it will need to be investigated before continuing.

Execution

Once you have confirmed that a dry-run will execute successfully, it is time to start a real run of the workflow. This is done much in the same way as a dry run.

1. Start your screen session: screen -S snakemake
2. Change directories to the workflow (where it was downloaded with git clone . . .). I don't know where you downloaded the workflow, so this part's up to you
3. Activate our conda environment: conda activate snakemake
4. Start the workflow with snakemake --profile slurm (or ~/.profile/snakemake/slurm if you receive errors at this step)
5. Once you have verified the workflow is starting with no errors, perform the following to safely detach the screen session (which keeps snakemake running)
   1. CTRL + a, d
      1. Press and hold the control key
      2. Press a. Continue holding control.
      3. Press d
      4. The session will exit. Verify with screen -ls. It should say (Detached) in the output.

From here, the workflow should continue running without errors. If you would like to examine the progress, simply attach back into the screen session.
After logging into the cluster, execute the following:

1. screen -r snakemake
2. Your snakemake session is active, with all previous output.
3. Exit the screen sesion with the steps as described above

Any log files will be found in the logs directory under the same folder snakemake was started in. Each rule has its own folder, with output files containing the information they are running on (tissue_name, run number, etc.)

Go back to Setup
Return to the Home Page