Snakeboost

Snakeboost provides enhancers and helpers to turbocharge your snakemake workflows. The project is currently in it’s alpha stage.

Script Enhancers

Script enhancer functions wrap around bash scripts given to the shell directive in Snakemake Rules. All enhancers have a common interface designed for easy use in your workflow. To illustrate, we’ll take PipEnv as an example (it lets you use pip virtual environments!).

Initiate the enhancer

Import the enhancer at the top of your Snakefile and instantiate it. Most enhancers take a few arguments defining their global settings.
```
from snakeboost import PipEnv

my_env = PipEnv(packages=["numpy", "flake8"], root=("/tmp"))
```
Use the enhancer in a rule

When instantiated, enhancers can be called using the bash command as an argument.
```
rule lint_python:
    inputs: "some-script.py"
    shell:
        my_env.script("flake8 {input}")
```
Some enhancers, such as PipEnv, provide multiple functions (e.g. .script, .python, etc) that provide slightly different functionality. Others, such as Tar, have methods that return a modified instance.
```
from snakeboost import Tar

tar = Tar(root="/tmp")
rule inspect_tarball:
    inputs: "some_archive.tar.gz"
    shell:
        tar.using(inputs=["{input}"])("ls {input}/")
```
Here, tar.using() returns a new instance of Tar with modified properties. Tar itself is callable, taking a string and wrapping it with all the script necessary to unpack your input tar file. Snakeboost uses this slightly convoluted way of setting arguments to allow easy chaining of multiple enhancers. This leads us to step 3:

Use boost() to chain multiple enhancers

Chaining many enhancers together can quickly lead to indentation hell:

rule lint_tarred_scripts:
    inputs: "script_archive.tar.gz"
    shell:
        xvfb-run(
            tar.using(inputs=["{input}"])(
                my_env.script(
                    "flake8 {input}/script.py"
                )
            )
        )

The boost() function lets you rewrite this as:

from snakeboost import Boost

boost = Boost()

rule lint_tarred_scripts:
    inputs: "script_archive.tar.gz"
    shell:
        boost(
            xvfb_run,
            tar.using(inputs=["{input}"]),
            my_env.script,
            "flake8 {input}/script.py"
        )

That makes your rules much cleaner! However, boost provides a much more important function, described in full detail on the boost page…

Contributing

If you have a small utility function for your Snakemake workflows, it would likely make a great addition to the Snakeboost ecosystem. Script enhancers should follow the basic interface of the other enhancers: a class initialized with global settings that exposes one or more functions that take a bash script as argument.

Snakebids uses Poetry for dependency management and pre-commit for Quality Assurance tests. If Poetry is not already installed on your system, follow the instructions on their website. Then clone the repo and initialize by running:

poetry install
poetry run pre-commit install