Step by step tutorial with the sample data set¶
Setup¶
Make sure you have Singularity and Snakemake up and running. We have experienced problems with singularity pull
and docker images when using a temporary directory (TMPDIR) in the shared file system. Setting TMPDIR=/tmp/
resolves this issue.
Installation¶
Follow Installation guide
Configuration¶
Open Baltica/data/config.yml
and replace /beegfs/homes/tbrittoborges/
or /home/tbrittoborges
to your desired path. The path parameter specifies where the project will be located. There is no needs to change baltica_path, as Baltica resolves it.
Execution¶
Execute Baltica with baltica all /Baltica/config.yml --use-singularity
. Add --quiet to reduce the logging level or --verbose to increase it. Use a cluster profile (--profile) to take advantage of the cluster scheduler. See more useful parameters for snakemake below.
You should expect a results/
directory containing the most relevant files by the end of the run. The report and excel table, results/baltica_report{project_title}.html
and results/baltica_table{project_title}.xlsx
, are the most relevant files. Intermediate files are kept in directories for each method (named for the methods), and can be used or deleted.
Important snakemake parameters¶
- --cores: only required if you are not using a cluster scheduler. Use
baltica ... --cores all
to specify the maximum number of cores available. - --profile: setup a cluster configuration profile.
- --dry-run: only computes the DAG, does not execute rules. But execution order may not reflect the actual order of execution.
- --unlock: unlock the directory if a previous run had problems.
- --list-untracked: list files that are not tracked by workflows, useful if you want to clean up the directory to save disk space after a successful run.
- --quiet: less verbose output.
- --reason, --printshellcmds, --verbose: verbose output.
Simple use baltica ... --verbose
to get maximum debug information (you may want to redirect it to a file).
See more detail at the Snakemake docs.