☰ Menu

      UC Davis Bioinformatics Workshop Base Template

Home
Introduction
Intro to the Workshop and Core
Examples of Markdown Formatting
Snakemake
Introduction
RNAseq/TagSeq Workflow
10x Supernova (In progress)

Running UCD Bioinformatics Supernova Pipeline with Snakemake

alt text

HOW TO CALL THE SNAKEMAKE FILE (GENERAL):

  1. snakemake -s snakemake.py **RULE_NAME**
  2. Files needed: snakemake.py, templates/cluster.json, templates/keith.json
  3. Be sure to specify running_locally in the template/keith.json file.
    • If running_locally == ‘True’:
      • templates/cluster.json is not needed if running locally only.
      • don’t need to worry about the sbatch parameters in the templates/keith.json file
      • Follow rules under RUNNING LOCALLY below.
    • Else if running_locally == ‘False’:
      • be sure to set all sbatch parameters in the templates/keith.json file
      • Follow rules under RUNNING ON CLUSTER below.

RUNNING ON CLUSTER

  1. Left Side
    • running_locally = “False”, sbatch included in snakemake call and run supernova runs on that job and call sbatch for mkoutput (3x)
    • snakemake -s snakemake.py --cluster-config templates/cluster.json --cluster "sbatch -J {cluster.ls.job-name} -N {cluster.ls.n} -p {cluster.ls.partition} -t {cluster.ls.time} -m {cluster.ls.mem} --output {cluster.ls.output} --error {cluster.ls.error} --mail-type {cluster.ls.mail-type} --mail-user {cluster.ls.mail-user} --exclusive" left_side
  2. Right Side
    • running_locally = “False”, sbatch included in snakemake call and proc10xg runs on that job and call sbatch for kat reads (3x)
    • snakemake -s snakemake.py --cluster-config templates/cluster.json --cluster "sbatch -J {cluster.rs.job-name} -N {cluster.rs.n} -p {cluster.rs.partition} -t {cluster.rs.time} -n {cluster.rs.ntasks} -m {cluster.rs.mem} --output {cluster.rs.output} --error {cluster.rs.error} --mail-type {cluster.rs.mail-type} --mail-user {cluster.rs.mail-user}" right_side
  3. Bottom
    • running_locally = “False”, no sbatch included in call or included (doesnt matter) all portions submitted as an sbatch
    • snakemake -s snakemake.py bottom
  4. Running 1 rule of interest?
    • running_locally = “False”,
    • All rules except run_supernova and proc10xG
      • snakemake -s snakemake.py **RULE_NAME**
    • run_supernova
      • Run number 1, except replace left_side with run_supernova at the end of the command.
    • proc10xG
      • Run number 2, except replace right_side with proc10xG at the end of the command.

dag

RUNNING LOCALLY (TODO)

  1. All
    • running_locally = “True”
    • snakemake -s snakemake.py all
  2. Left Side
    • running_locally = “True”
    • snakemake -s snakemake.py all
  3. Right Side
    • running_locally = “True”
    • snakemake -s snakemake.py all
  4. Bottom
    • running_locally = “True”
    • snakemake -s snakemake.py all
  5. Running 1 rule of interest? (IDK ABOUT THIS QUITE YET)
    • running_locally = “True”, no sbatch included in call
    • snakemake -s snakemake.py **RULE_NAME** ***

TODO pick better names then Left and Right side since it changes when you make the graph …haha