aamutils 0.0.3

Collection of atom-atom-mapping utility functions.

Atom-atom-mapping Utils

A collection of atom-atom-mapping utility functions.

Installation

The easiest way to use AAMUtils is by installing the PyPI package aamutils.

pip install aamutils

Usage

The input is a list of partial atom-atom-maps (AAMs). Data is read line-by-line from a text file. Each line should contain one reaction SMILES.

Here is a simple example extending the partial AAM to a complete AAM. First generate the input data:

echo "CCC[Cl:1].[N:2]>>CCC[N:2].[Cl:1]" > testinput.txt

Next, run AAMUtils to expand the partial AAM.

python3 -m aamutils expand testinput.txt

The output is written to 'testinput_extended.json'.

cat testinput_extended.json 

[
    {
        "input": "CCC[Cl:1].[N:2]>>CCC[N:2].[Cl:1]",
        "expanded_aam": "[Cl:1][CH2:5][CH2:4][CH3:3].[NH3:2]>>[ClH:1].[NH2:2][CH2:3][CH2:4][CH3:5]",
        "ilp_status": "Optimal",
        "optimization_result": 4.0,
        "invalid_reaction_center": false,
        "reaction_edges": 4
    }
]

Benchmark

To rerun the benchmarks from the paper use the benchmark.py script. The reported results can be reproduced by running the following commands:

(1) Extend partial reaction center (50%, 75% and 100% missing atoms)

python3 benchmark.py --remove-mode rc --remove-ratio 0.5

python3 benchmark.py --remove-mode rc --remove-ratio 0.75

python3 benchmark.py --remove-mode rc --remove-ratio 1

(2) Extend partial AAM with fully mapped RC

python3 benchmark.py --remove-mode keep_rc --remove-ratio 1

Functionality

Here is an overview of implemented functionality:

• SMILES to graph and graph to SMILES parsing
• Reaction center validity checks
• ITS graph generation
• Expand partial AAM to complete AAM on balanced reactions
• AAMing based on minimal chemical distance (MCD) for balanced reactions

License

This project is licensed under MIT License - see the License file for details.

Acknowledgments

This project has received funding from the European Unions Horizon Europe Doctoral Network programme under the Marie-Skłodowska-Curie grant agreement No 101072930 (TACsy -- Training Alliance for Computational)