Introduction

With Chemoton you can explore complex chemical reaction networks in an automated fashion. Based on a Python framework, workflows can be built that probe reactivity of chemical systems with quantum chemical methods. Various quantum chemical software programs and job schedulers are supported via the back-end software SCINE Puffin.

Installation

Prerequisites

The key requirements for Chemoton are the Python packages scine_utilities, scine_database, and scine_molassember. These packages are available from PyPI and can be installed using pip. However, these packages can also be compiled by hand. For the latter case please visit the repositories of each of the packages and follow their guidelines.

Installation

Chemoton can be installed using pip (pip3) once the repository has been cloned:

git clone <chemoton-repo> chemoton
cd chemoton
pip install -r requirements.txt
pip install .

A non-root user can install the package using a virtual environment, or the --user flag.

The documentation can be found online, or it can be built using:

make -C docs html

It is then available at:

<browser-name> docs/build/html/index.html

In order to build the documentation, you need a few extra Python packages, which are not installed automatically together with Chemoton. In order to install them, run

pip install -r requirements-dev.txt

Tutorial

Minimal Example

Assuming that Chemoton has successfully been installed, a small example exploration can be started by running Chemoton’s main function. It requires a database running on localhost listening to the default mongodb port 27017; additionally a puffin instance has to be running and checking the database named default.

Setting up these things may look somewhat like this:

1. Start a mongodb server. Limit its memory usage and maybe customize where to log and store the data.

mongod --fork --port=27017 -dbpath=<path to db storage dir> -wiredTigerCacheSizeGB=1 --logpath=mongo.log

  1. Configure and bootstrap a puffin:

python3 -m puffin configure
# Edit the generated puffin.yaml here
python3 -m puffin -c puffin.yaml bootstrap

3. Source the puffin settings and tell it to listen to the correct DB. (Hostname and port should be the default ones.) Then start it.

source puffin.sh
export PUFFIN_DATABASE_NAME=default
python3 -m puffin -c puffin.yaml start

  1. Run the Chemoton exploration defined in the __main__ function:

python3 -m scine_chemoton wipe

The optional wipe argument will start the example exploration with a clean puffin_tests DB; giving the continue argument will reuse old data.

Expanding on the Minimal Example

The functionalities used in Chemoton’s __main__.py are a good starting point for most simple explorations. The file contains a lot of settings that are explicitly set to their defaults in order to show their existence.

While we recommend to read the documentation of Chemoton, tinkering with explorations can be as simple as:

cp <chemoton-git>/scine_chemoton/__main__.py my_awesome_exploration.py

and editing the file to your liking: disabling gears, adding filters or just changing methods.

How to Cite

When publishing results obtained with Chemoton, please cite the corresponding release as archived on Zenodo (DOI 10.5281/zenodo.6695583; please use the DOI of the respective release).

In addition, we kindly request you to cite the following article when using Chemoton:

J. P. Unsleber, S. A. Grimmel, M. Reiher, “Chemoton 2.0: Autonomous Exploration of Chemical Reaction Networks”, arXiv:2202.13011 [physics.chem-ph].

Support and Contact

In case you should encounter problems or bugs, please write a short message to scine@phys.chem.ethz.ch.