Conda How-To

Introduction to Conda

Conda is a package manager that automates the installation of scripts, packages and dependencies in a user-friendly way. It uses the concept of environments that are independent and don’t interact with each other, limiting the possibility of dependency or package conflicts.

A usual conda workflow looks like this:

Once a conda environment was created it can be activated or de-activated as many times as you want. The tools etc. installed inside the environment won’t disappear, e.g., when you re-start the computer or log out.

Create a new Conda environment

To create a new environment use the command conda create plus the –name flag to set a name for your environment.

Let’s create a new environment called ont_test.

course_user> conda create --name ont_test

You should see the following output …

course_user> conda create --name ont_test
Collecting package metadata (current_repodata.json): done
Solving environment: done

… and after a little while conda will let you know the path it will try to install the new environment to and ask you to proceed. Simply press Enter …

## Package Plan ##

environment location: /home/course_user/miniconda3/envs/ont_test



Proceed ([y]/n)? y

Preparing transaction: done
Verifying transaction: done
Executing transaction: done
#
# To activate this environment, use
#
#     $ conda activate ont_test
#
# To deactivate an active environment, use
#
#     $ conda deactivate

Retrieving notices: ...working... done

Finished. We now have a new conda environment called ont_test.

To list all conda environments use the command conda env list. If the only environment on your system was the one we just installed you would see an output similar to this:

# conda environments:
#
base                  *  /home/course_user/miniconda3
ont_test              *  /home/course_user/miniconda3/ont_test

There it is, in /home/course_user/miniconda3/ont_test. However, on your virtual machine you will see several other environments in addition to the one you just created, but ignore these for the moment.

To remove the environment, all one would have to do is to remove the directory using rm -r

course_user> rm -r /home/course_user/miniconda3/ont_test

If we now list the existing conda environments the only one left would be Conda’s base environment

course_user> conda env list

# conda environments:
#
base                  *  /home/course_user/miniconda3
If you just removed the ont_test environment to check that it really disappears then please re-create it again. We will need it in the next section.

Activate a conda environment

Now that we have created our new environment we want to be able to use it. To use a conda environment one has to activate it, using the conda activate command followed by the name of the environment:

course_user> conda activate ont_test
(ont_test) course_user>

As you can see, after activating the ont_test environment, the command-line adds the name of the environment in brackets to our prompt. This shows us that we’re inside the conda environment and can start to use it.

To get out of the environment we can use the conda command deactivate

(ont_test) course_user> conda deactivate
course_user>

Installing packages using conda

Now that we set up a new clean conda environment we need to install something in it.

Conda packages can be installed from various so called channels. For example, the channel bioconda provides packages related to the biomedical research space including packages and tools for Oxford Nanopore data processing. One of these tools is porechop, a tool to find and remove sequencing adapters from Oxford Nanopore reads.

To install something using conda we use the command conda install PACKAGE_NAME where PACKAGE_NAME is the name of the tool/package we want to install. To specify a particular channel the tool can be found on we use the -c flag followed by the channel name.

Let’s activate our ont_test environment and install the tool porechop from channel bioconda in it:

course_user> conda activate ont_test
(ont_test) course_user> conda install -c bioconda porechop

Once you’ve hit Enter you will see the Collecting package metadata and Solving environment messages of conda, followed by a list of packages and dependencies conda will install. Hit Enter again to proceed with the installation process.

Collecting package metadata (current_repodata.json): done
Solving environment: done

## Package Plan ##

  environment location: /home/course_user/miniconda3/envs/ont_test

  added / updated specs:
    - porechop


The following packages will be downloaded:

    package                    |            build
    ---------------------------|-----------------
    porechop-0.2.4             |  py310h30d9df9_3         105 KB  bioconda
    python-3.10.6              |ha86cf86_0_cpython        29.0 MB  conda-forge
    ------------------------------------------------------------
                                       Total:        29.1 MB

The following NEW packages will be INSTALLED:

  _libgcc_mutex      conda-forge/linux-64::_libgcc_mutex-0.1-conda_forge None
  _openmp_mutex      conda-forge/linux-64::_openmp_mutex-4.5-2_gnu None
  bzip2              conda-forge/linux-64::bzip2-1.0.8-h7f98852_4 None
  ca-certificates    conda-forge/linux-64::ca-certificates-2022.9.24-ha878542_0 None
  ld_impl_linux-64   conda-forge/linux-64::ld_impl_linux-64-2.39-hc81fddc_0 None
  libffi             conda-forge/linux-64::libffi-3.4.2-h7f98852_5 None
  libgcc-ng          conda-forge/linux-64::libgcc-ng-12.2.0-h65d4601_19 None
  libgomp            conda-forge/linux-64::libgomp-12.2.0-h65d4601_19 None
  libnsl             conda-forge/linux-64::libnsl-2.0.0-h7f98852_0 None
  libsqlite          conda-forge/linux-64::libsqlite-3.39.4-h753d276_0 None
  libstdcxx-ng       conda-forge/linux-64::libstdcxx-ng-12.2.0-h46fd767_19 None
  libuuid            conda-forge/linux-64::libuuid-2.32.1-h7f98852_1000 None
  libzlib            conda-forge/linux-64::libzlib-1.2.13-h166bdaf_4 None
  ncurses            conda-forge/linux-64::ncurses-6.3-h27087fc_1 None
  openssl            conda-forge/linux-64::openssl-3.0.7-h166bdaf_0 None
  pip                conda-forge/noarch::pip-22.3.1-pyhd8ed1ab_0 None
  porechop           bioconda/linux-64::porechop-0.2.4-py310h30d9df9_3 None
  python             conda-forge/linux-64::python-3.10.6-ha86cf86_0_cpython None
  python_abi         conda-forge/linux-64::python_abi-3.10-2_cp310 None
  readline           conda-forge/linux-64::readline-8.1.2-h0f457ee_0 None
  setuptools         conda-forge/noarch::setuptools-65.5.1-pyhd8ed1ab_0 None
  tk                 conda-forge/linux-64::tk-8.6.12-h27826a3_0 None
  tzdata             conda-forge/noarch::tzdata-2022f-h191b570_0 None
  wheel              conda-forge/noarch::wheel-0.38.4-pyhd8ed1ab_0 None
  xz                 conda-forge/linux-64::xz-5.2.6-h166bdaf_0 None


Proceed ([y]/n)? 



Downloading and Extracting Packages
porechop-0.2.4       | 105 KB    | #################################################################################################### | 100% 
python-3.10.6        | 29.0 MB   | #################################################################################################### | 100% 
Preparing transaction: done
Verifying transaction: done
Executing transaction: done
Retrieving notices: ...working... done

Done! Now porechop is installed in your conda environment. To check that everything went well check the porechop version

(ont_test) course_user> porechop --version
0.2.0

Creating a conda environment with pre-installed packages

Now we know how to create a conda environment, activate it and install packages and tools in it. However, we can also install packages in environments at the time of creation. We can even set a specific version of the packages and tools we want to install.

The following command will

course_user> conda create --name ont_test2 python==3.6 porechop

Once Conda is done with the installation activate the environment ont_test2 and check the version of python and porechop

course_user> conda activate ont_test2
(ont_test2) course_user> python --version
3.6.0
(ont_test2) course_user> porechop --version
0.2.4

Tip
You can also install packages in already existing environments without activating it using the conda install command but specifying the name of the environment you want to install the package into using the --name flag:
conda install --name ENV_NAME PACKAGE_NAME

Great. Now you have a good basis to work with Conda environments

Sharing Conda environments

Conda environments can also be shared. To do so one has to create a specific file, an environment.yaml file outlining the name of the environment, dependencies and tools to be installed.

For example the environment.yaml file for this course looks something like this:

name: biosec_course
channels: 
  - bioconda
dependencies:
  - minimap
  - miniasm
  - flye
  - assembly-stats
...

To create a conda environment from scratch one would have to specify the environment.yaml with the -f flag, e.g.:

course_user> conda env create -f environment.yaml

Conda Self-Check

  1. How do you list all conda environments already installed?
  2. How would you activate a the conda environment anu_course?
  3. What command would you use if you wanted to install the tool jellybean from channel conda-forge?
  4. If you deactivate the environment ont_test2 and check the version of porechop, what will you see? Why?

Answers

For a short Conda Cheat-Sheet click here.

CONTINUE ->