SNAKEMAKE at CNAG

Set up enviroment

Get snakemake >= 6.0.5 and python3 in your environment. This script has been tested with Snakemake v6.0.5

Set up your pipeline

Currently sopported:

• threads
• time
• name
• qos
• partition
• constraint
• array

These can be set up in snakefile or cluster-config file.

snakefile

rule example1:
    input: {input}
    output: {output}
    threads: 1
    params:
        time = "10:00:00"
        name = "JobName" 

cluster-config.json

{
    "example1":
        {
            "name": "example1_job",
            "time": "23:00:00",
            "threads": 2
        }
}

If one of the rules uses wildcards, you can specify a different jobname per wildcard (eg. using the wildcard "file"):

{
    "example1":
        {
            "name": "example1_job.{wildcards.file}",
            "time": "23:00:00",
            "threads": 2
        }
}

You must take into account, config-cluster file overwrites snakefile configuration.

Running a pipeline with dependencies at CNAG's cluster

snakemake --jobs 999 --is --nt [...] \
--snakefile "mySnakeFile" \
--cluster "sbatch-cnag.py {dependencies}" \
--config "pipeline_config.json" \
--cluster-config "cluster-config.json"

Writing your pipelines

Full snakemake documentation can be found here