This is a place to collect the useful material for my Computational Biology course

Environments: Choose the right room to accomplish a task

More often than not, to accomplish a task you need several tools (i.e. software), and according to the task itself, you need specific tools.

Solving different tasks will lead you to continuously add new software that increase the probability of software conflicts.

Creating an environment for each task (or small group of heterogeneous tasks) will help to avoid software conflicts.

conda is the tool to easily manage environments.

Installing mamba

First of all you need to install the package manager. I suggest to use miniforge.

• wget "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh" bash Miniforge3-$(uname)-$(uname -m).sh
• miniforge3/bin/mamba init

As an alternative you can use the installer tailored for the Virtual Machines @UniMiB:

wget https://raw.githubusercontent.com/qLSLab/CBCourse/main/VMconfigs/installMyMamba.sh

and then execute it from the home

Creating an environment

To create a new environment is simple as this:

mamba create -n MyEnvironmentName

then you need to activate it

mamba activate MyEnvironmentName

and populate it with the available software of your choice

mamba install python=3.11 pandas matplotlib

With the line above, you will install python (version 3.11), its two libraries matplotlib and pandas, with all the necessary programs (dependencies).

From now on, whenever you need to use python together with pandas or matplotlib, it is enough to activate de environment and use them.

Remember that the programs installed within an environment are available only within that environment.

Using an environment

To "enter" an environment you need to explicitly activate it:

mamba activate environmentName

Once finished, you can "exit" the environment by deactivating it:

mamba deactivate

To obtain the list of all the available environments, digit

mamba info --envs

Sharing an environment

It is possible to save a snapshot of an environment in order to share it, or to port it on another computer.

mamba activate MyEnvironmentName
mamba env export > environment.yml

then, you can create an exact copy of it running

conda env create -f environment.yml

or

mamba env update -n <your-env> --file environment.yml

Removing an environment

To completely remove an environment from your computer, digit

mamba env remove --name MyEnvironmentName

Conda channels

A conda/mamba channel is a compilation of packages. Some of them are thematics such as bioconda or tries to have the most up to date cross platform version of the packages conda-forge

To add a channel to an environment run

mamba config --append channels conda-forge

Example:

Create an environment to perform exploratory data analyses with python. Let's say we need python (3.11), ipython (to ease the use of python), pandas (to mangle the data), openpyxl (to write *.xlsx) and matplotlib to visualize the information.

mamba create -n dataAn python=3.11
mamba activate dataAn
mamba install ipython pandas openpyxl matplotlib

or from the file condaEnvDataAnalysisMinimal.yml

conda create -f condaEnvDataAnalysisMinimal.yml

mamba and poetry

1. If not already done, install miniforge or micromamba:

   • wget "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh" bash Miniforge3-$(uname)-$(uname -m).sh
   • miniforge3/bin/mamba init

2. Create a new specific environment for your project:

   mamba create -n myNewProjectEnv python=3.XX poetry="1.*"

3. activate it: mamba activate myNewProjectEnv

4. create a new project poetry new myNewProject or tune the pyproject.toml

5. Install all the dependencies poetry install

6. run your code either from mamba or poetry e.g poetry run ipython myProgram.py

Visual Studio Code tips:

• to render invisible charaters "editor.renderWhitespace": "all", (ctrl, -> search whitespace)
• Select and activate an environment