Accurate understanding of Ultraviolet–visible (UV–Vis) spectra is critical for highthroughput design of compounds for drug discovery. Experimentally determining UV–Vis spectra can become expensive when dealing with a large quantity of novel molecules. This provides us an opportunity to drive computational advances in molecular property predictions using quantum mechanics and machine learning. In this work, we use both Quantum Mechanically (QM) predicted and measured UV–Vis spectra as input to modify four different machine learning architectures: UVvis-SchNet, UVvis- DTNN, UVvis-Transformer, and UVvis-MPNN. Here we find that the UVvis-MPNN model outperforms the other models when using optimized 3D coordinates and QM predicted spectra as input features. This model has the highest performance for predicting UVVisible spectra with a training RMSE of 0.06 and validation RMSE of 0.08.
McNaughton, Andrew D., et al. "Machine Learning Models for Predicting Molecular UV–Vis Spectra with Quantum Mechanical Properties." Journal of Chemical Information and Modeling, vol. 63, no. 5, Feb. 2023. https://doi.org/10.1021/acs.jcim.2c01662
McNaughton, Andrew D., Joshi, Rajendra Prashad, Knutson, Carter R., Fnu, Anubhav, Luebke, Kevin J., Malerich, Jeremiah P., Madrid, Peter B., & Kumar, Neeraj (2023). Machine Learning Models for Predicting Molecular UV–Vis Spectra with Quantum Mechanical Properties. Journal of Chemical Information and Modeling, 63(5). https://doi.org/10.1021/acs.jcim.2c01662
McNaughton, Andrew D., Joshi, Rajendra Prashad, Knutson, Carter R., et al., "Machine Learning Models for Predicting Molecular UV–Vis Spectra with Quantum Mechanical Properties," Journal of Chemical Information and Modeling 63, no. 5 (2023), https://doi.org/10.1021/acs.jcim.2c01662
@article{osti_1964168,
author = {McNaughton, Andrew D. and Joshi, Rajendra Prashad and Knutson, Carter R. and Fnu, Anubhav and Luebke, Kevin J. and Malerich, Jeremiah P. and Madrid, Peter B. and Kumar, Neeraj},
title = {Machine Learning Models for Predicting Molecular UV–Vis Spectra with Quantum Mechanical Properties},
annote = {Accurate understanding of Ultraviolet–visible (UV–Vis) spectra is critical for highthroughput design of compounds for drug discovery. Experimentally determining UV–Vis spectra can become expensive when dealing with a large quantity of novel molecules. This provides us an opportunity to drive computational advances in molecular property predictions using quantum mechanics and machine learning. In this work, we use both Quantum Mechanically (QM) predicted and measured UV–Vis spectra as input to modify four different machine learning architectures: UVvis-SchNet, UVvis- DTNN, UVvis-Transformer, and UVvis-MPNN. Here we find that the UVvis-MPNN model outperforms the other models when using optimized 3D coordinates and QM predicted spectra as input features. This model has the highest performance for predicting UVVisible spectra with a training RMSE of 0.06 and validation RMSE of 0.08.},
doi = {10.1021/acs.jcim.2c01662},
url = {https://www.osti.gov/biblio/1964168},
journal = {Journal of Chemical Information and Modeling},
issn = {ISSN 1549-9596},
number = {5},
volume = {63},
place = {United States},
publisher = {American Chemical Society},
year = {2023},
month = {02}}