Improved environmental chemistry property prediction of molecules with graph machine learning

Zhu, Shang; Nguyen, Bichlien H.; Xia, Yingce; Frost, Kali; Xie, Shufang; Viswanathan, Venkatasubramanian; Smith, Jake A.

doi:10.1039/d3gc01920a

Cited by 4 publications

(2 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Despite certain challenges in model building and training for ML, such as data availability, model interpretability, and overfitting, the amalgamation of ML with computational chemistry holds immense potential as a catalyst for advancements in material molecular design, drug discovery, and other domains. 168–171…”

Section: Machine Learning-based Performance Predictionmentioning

confidence: 99%

Optimizing the performance of phase-change azobenzene: from trial and error to machine learning

Wang,

Yu,

Gao

et al. 2024

J. Mater. Chem. C

View full text Add to dashboard Cite

show abstract

Section: Machine Learning-based Performance Predictionmentioning

confidence: 99%

Optimizing the performance of phase-change azobenzene: from trial and error to machine learning

Wang,

Yu,

Gao

et al. 2024

J. Mater. Chem. C

View full text Add to dashboard Cite

show abstract

“…Among many AI approaches, Graph Neural Networks (GNNs) have been used for such tasks, due to their ability to encode the inherent graph-like nature of molecules, enabling the incorporation of spatial and contextual information [13][14][15][16] . An alternative method for representing molecular structures is a string-based SMILES notation, which can serve as a direct input to Large Language Models (LLMs) 8,17 .…”

mentioning

confidence: 99%

ChemBERTa-2: Fine-Tuning for Molecule’s HIV Replication Inhibition Prediction

Nowakowska

2023

Preprint

View full text Add to dashboard Cite

show abstract

Graph Neural Network-Based Molecular Property Prediction with Patch Aggregation

See,

Zhang,

Boley

et al. 2024

J. Chem. Theory Comput.

View full text Add to dashboard Cite

Graph neural networks (GNNs) have emerged as powerful tools for quantum chemical property prediction, leveraging the inherent graph structure of molecular systems. GNNs depend on an edge-to-node aggregation mechanism for combining edge representations into node representations. Unfortunately, existing learnable edge-to-node aggregation methods substantially increase the number of parameters and, thus, the computational cost relative to simple sum aggregation. Worse, as we report here, they often fail to improve predictive accuracy. We therefore propose a novel learnable edge-to-node aggregation mechanism that aims to improve the accuracy and parameter efficiency of GNNs in predicting molecular properties. The new mechanism, called "patch aggregation", is inspired by the Multi-Head Attention and Mixture of Experts machine learning techniques. We have incorporated the patch aggregation method into the specialized, state-of-the-art GNN models SchNet, DimeNet++, SphereNet, TensorNet, and VisNet and show that patch aggregation consistently outperforms existing learnable and nonlearnable aggregation techniques (sum, multilayer perceptron, softmax, and set transformer aggregation) in the prediction of molecular properties such as QM9 thermodynamic properties and MD17 molecular dynamics trajectory energies and forces. We also find that patch aggregation not only improves prediction accuracy but also is parameter-efficient, making it an attractive option for practical applications for which computational resources are limited. Further, we show that Patch aggregation can be applied across different GNN models. Overall, Patch aggregation is a powerful edge-to-node aggregation mechanism that improves the accuracy of molecular property predictions by GNNs.

show abstract

Improved environmental chemistry property prediction of molecules with graph machine learning

Abstract: Rapid prediction of environmental chemistry properties is critical towards the green and sustainable development of chemical industry and drug discovery. Machine learning methods can be applied to learn the relations...

Cited by 4 publications

References 27 publications

Optimizing the performance of phase-change azobenzene: from trial and error to machine learning

Optimizing the performance of phase-change azobenzene: from trial and error to machine learning

ChemBERTa-2: Fine-Tuning for Molecule’s HIV Replication Inhibition Prediction

Graph Neural Network-Based Molecular Property Prediction with Patch Aggregation

Contact Info

Product

Resources

About