DOE CODE   /    Search Results   /    pnnl/emp-gnn

pnnl/emp-gnn

RESOURCE

Publicly Accessible Repository
https://github.com/pnnl/emp-gnn
https://doi.org/10.11578/dc.20240229.1

SAVE / SHARE



PNNL Badge

Abstract

Efficient Graph Neural Network for Predicting Molecular Properties software can compare the prediction quality with ab initio DFT results reported by the high-performance state-of-theart NWChem quantum chemistry package [1] through Mean Absolute Error (MAE) obtained by the fitting between DFT and model predictions i.e, MPNN [2]. We also demonstrated the performance benefits by grouping large molecules by atom sizes, and executing GNN models on different types of resources. Since the training times depend on the number of atoms, we demonstrate the impact of distributing the workloads on two GPUs with varying capabilities (e.g., NVIDIA A100 vs. GeForce RTX 2080 Ti) to optimize the efficiency. We are at the precipice of broad adoption of GNNs for molecular property prediction tasks; hence, our work is timely by comparing model prediction against classical approaches with the intent of providing initial screening for specific classes of molecules.
Developers:
Lee, Hyungro [1] ;  Ghosh, Sayan [2]
  1. PNNL
  2. Pacific Northwest National Laboratory (PNNL), Richland, WA (United States)
Release Date:
2024-02-29
Project Type:
Open Source, Publicly Available Repository
Software Type:
Scientific
Licenses:
BSD 2-clause "Simplified" License
Sponsoring Org.:
Code ID:
123162
Site Accession Number:
Battelle IPID 32882-E
Research Org.:
Pacific Northwest National Laboratory (PNNL), Richland, WA (United States)
Country of Origin:
United States

RESOURCE

Publicly Accessible Repository
https://github.com/pnnl/emp-gnn
https://doi.org/10.11578/dc.20240229.1

SAVE / SHARE



PNNL Badge

Citation Formats

Lee, Hyungro, and Ghosh, Sayan. pnnl/emp-gnn. Computer Software. https://github.com/pnnl/emp-gnn. USDOE. 29 Feb. 2024. Web. doi:10.11578/dc.20240229.1.
Lee, Hyungro, & Ghosh, Sayan. (2024, February 29). pnnl/emp-gnn. [Computer software]. https://github.com/pnnl/emp-gnn. https://doi.org/10.11578/dc.20240229.1.
Lee, Hyungro, and Ghosh, Sayan. "pnnl/emp-gnn." Computer software. February 29, 2024. https://github.com/pnnl/emp-gnn. https://doi.org/10.11578/dc.20240229.1.
@misc{ doecode_123162,
title = {pnnl/emp-gnn},
author = {Lee, Hyungro and Ghosh, Sayan},
abstractNote = {Efficient Graph Neural Network for Predicting Molecular Properties software can compare the prediction quality with ab initio DFT results reported by the high-performance state-of-theart NWChem quantum chemistry package [1] through Mean Absolute Error (MAE) obtained by the fitting between DFT and model predictions i.e, MPNN [2]. We also demonstrated the performance benefits by grouping large molecules by atom sizes, and executing GNN models on different types of resources. Since the training times depend on the number of atoms, we demonstrate the impact of distributing the workloads on two GPUs with varying capabilities (e.g., NVIDIA A100 vs. GeForce RTX 2080 Ti) to optimize the efficiency. We are at the precipice of broad adoption of GNNs for molecular property prediction tasks; hence, our work is timely by comparing model prediction against classical approaches with the intent of providing initial screening for specific classes of molecules.},
doi = {10.11578/dc.20240229.1},
url = {https://doi.org/10.11578/dc.20240229.1},
howpublished = {[Computer Software] \url{https://doi.org/10.11578/dc.20240229.1}},
year = {2024},
month = {feb}
}