GitHubHelp

misol/ml-D12-water-app

Prediction of binary diffusivities of any solute in water at atmospheric pressure

★ 0Forks 0PythonGitHub ↗Compare

README

ml-D12-water-app

Prediction of binary diffusivities at infinite dilution of any solute in water at atmospheric pressure.

For more information see our paper: Aniceto et al., 2024, Journal of Molecular Liquids

Available models

The models in the Table below are ranked from better performance to worse. The AARD achieved by the models in the test set ranges from 3.92 to 5.48 %.

Model Input parameters required
ML-T-RDKit temperature and solute identifier (SMILES)
ML-T-Mordred temperature and solute identifier (SMILES)
ML-PropRD4it-4 temperature, solvent viscosoty, solute critical volume, molar refractivity (MolMR, see note)
ML-T-ECFP6 temperature and solute identifier (SMILES)
ML-Prop-4 temperature, solvent viscosoty, solute critical volume, solute Lenard-Jones diameter
ML-T-MACCS temperature and solute identifier (SMILES)

Note: the molar refractivity parameter (MolMR) can be computed from a molecular identifier (SMILES).

Installation

  1. Download this repository by clicking Code > Download ZIP. Unzip the folder.

  2. Install Python. We recommend a installing the Anaconda Distribution or Miniconda.

  3. Open the Anaconda Prompt and change the directory to where you extracted the repository files: cd path/to/folder.

  4. Create a conda virtual environment using the provided environment.yml file: conda env create -f environment.yml

  5. Activate the environment with: conda activate ml-water.

  6. You can now use the models following instructions bellow either in a .py script file or in a Jupyter Notebook (already provided in the environment by running jupyter lab).

Usage

To use the ML-Prop-4 model:

from ml_D12_water import ML_Prop_4

model = ML_Prop_4()

model.predict(temp=313, visc=0.6548, crit_vol=93.9, diam_lj=3.26)
# Output: array([2.77346518e-05])

To use the ML-T-RDKit model for two conditions/solutes:

from ml_D12_water import ML_T_RDKit

model = ML_T_RDKit()

model.predict(
    temp=[313, 303.2], 
    smiles=['C(=O)=O', 'O=CC1=CC=C(O)C(OC)=C1']  # CO2 and vanilin
)
# Output: array([2.81059139e-05, 1.01639130e-05])

The outputed D12 values are in cm2/s.

More usage examples are provided in the examples.ipynb file.

Citing

If these models are useful to you, please cite the following publication:

J.P.S. Aniceto, B. Zêzere, and C.M. Silva, Prediction of diffusion coefficients in aqueous systems by machine learning models, Journal of Molecular Liquids, 405, 125009, 2024, DOI: 10.1016/j.molliq.2024.125009.

Contributors

jAnicetomisol

Issues