A Python package for extracting and scoring reaction templates based on extended Hasse diagrams. The corresponding manuscript is available here.
Documentation: Documentation and a tutorial of EHreact is available at https://hesther.github.io/ehreact/.
Ehreact uses Python 3.6+ and a number of python package. We assume you have Conda on your system. If this is not the case yet, install Miniconda. However, Conda is optional and you may choose to install packages otherwise.
To install ehreact, make a new conda environment via
conda env create -f environment.ymlconda activate ehreactpip install -e .
(Note: If you have an old version of conda (< conda 4.4), the second command is source activate ehreact). The environment
simply contains python, numpy, rdkit, and typed-argument-parser. If you do not like to use conda, you can instead install these
packages yourself, and then run pip install -e ..
To check your installation, try to import ehreact in a Python session::
import ehreact
If you want to automatically atom-map reactions, download ReactionDecoder from Github (tested with rdt-2.4.1-jar-with-dependencies.jar), put the jar file into the folder ReactionDecoder (within the ehreact package folder) and rename the file to ReactionDecoder.jar This is optional, and you can run EHreact without ReactionDecoder.
Plotting Hasse diagrams relies on dot (install for example graphviz) and rsvg-convert. If you do not have these packages on your system already, you can install both via homebrew on Macos::
brew install graphvizbrew install librsvg
or via apt-get on Linux (Ubuntu, Debian)::
sudo apt-get install graphvizsudo apt-get install librsvg2-bin
There are other options to install, too. Installing dot and rsvg-convert is optional, and you can run EHreact without both, but will not be able to plot diagrams to an image.
Use train.py to create a Hasse diagram and specify --data_path <path> (the path to the known reactions or substrates), --train_mode <str> (transition_state for reaction mode if the input is reactions or single_reactant for single reactant mode if the input is single substrates), --save_path <path> (file to save the created diagram to, optional), and --save_plot <path> (file to save an image of the diagram, optional). For example, to create a diagram for the reactions in ehreact/data/reaction_training.dat, run
python train.py --data_path ehreact/data/reaction_training.dat --save_path test_transition_state.pkl --train_mode transition_state --save_plot test.png
which creates an image of the diagram saved under test.png and a pickle file of the diagram under test_transition_state.pkl. Please refer to the tutorial for more examples and options.
Use predict.py to score a set of query substrates or reactions. Specify --test_path <path> (the path to the query reactions or substrates), --load_path <path> (the path to a saved Hasse diagram pickle file), --predict_mode <str> (transition_state to score reactions, multi_reactant to propose products based on reactants and score all reactions, single_reactant to propose products and co-substrates based on a reactant and score all reactions). For example, to score the reactions in ehreact/data/reaction_test.dat, run
python predict.py --test_path ehreact/data/reaction_test.dat --load_path test_transition_state.pkl --predict_mode transition_state
Please refer to the tutorial for more examples and options.
Feel free to post questions, feedback, bugs or suggestions on github, or email to eheid@mit.edu.
Copyright (c) 2021, Esther Heid
Project based on the Computational Molecular Science Python Cookiecutter version 1.5.