mluvspec

Supplement files of paper "UV-Visible Absorption Spectra of Solvated Molecules by Quantum Chemical Machine Learning" .

Full text is also available in research gate and can be accessed here

This repository is structured as follow:

• machine-learning-python-files:

  • develop_ml_models.py: Develop 7-molecule (transfer-learning) models using cross-validation scheme
  • sample_output.log: Example output from running develop_ml_models.py with random state of 0
  • transfer_learn_with_pre_train_models.py: Transfer-learning for the target molecules with pre-trained models
  • utility.py: Helper functions for develop_ml_models.py and transfer_learn_with_pre_train_models.py

• simulation-files:

  • bc_calculations: Molecular geometries and FitSNAP input files used to calculate bispectrum components (BC)
  • fpmd: Input files used to perform first-principle molecular dynamics (fpmd), using cp2k
  • spectra_calculations: Input files for spectra calculations, using ORCA 4.2.1

• training-set-and-pre-train-model:

  • descriptors.tar.gz: BC descriptors computed for the 7-molecule (transfer-learning) model
  • pre_train_models.tar.gz: Pre-trained 7-molecule (transfer-learning) models
  • target_energies.tar.gz: 1st excited-state energies calculated with ORCA 4.2.1

How to cite

@article{doi:10.1021/acs.jctc.1c01181,
author = {Chen, Zekun and Bononi, Fernanda C. and Sievers, Charles A. and Kong, Wang-Yeuk and Donadio, Davide},
title = {UV–Visible Absorption Spectra of Solvated Molecules by Quantum Chemical Machine Learning},
journal = {Journal of Chemical Theory and Computation},
volume = {18},
number = {8},
pages = {4891-4902},
year = {2022},
doi = {10.1021/acs.jctc.1c01181},
    note ={PMID: 35913220},

URL = { 
        https://doi.org/10.1021/acs.jctc.1c01181
    
},
eprint = { 
        https://doi.org/10.1021/acs.jctc.1c01181
    
}

}