GraphDeep-hERG: Graph Neural Network PharmacoAnalytics for Assessing hERG-Related Cardiotoxicity

Yankang Jing^{1

2}, Yiyang Zhang^{1

2}, Guangyi Zhao^{1

2}, Terence McGuire^{1

2}, Jack Zhao^{1

2}, Ben Gibbs^{1

2}, Ganqian Hou^{1

2}, Zhiwei Feng^{3

4}, Ying Xue^{5

6}, Xiang-Qun Xie^{7

8

9}

Affiliations

¹ Department of Pharmaceutical Sciences and Computational Chemical Genomics Screening Center, Pharmacometrics & Systems Pharmacology (PSP) Pharmacoanalytics, School of Pharmacy, University of Pittsburgh, 6411 Salk Hall, 3501 Terrace Street, Pittsburgh, PA, 15261, USA.
² National Center of Excellence for Computational Drug Abuse Research, University of Pittsburgh, Pittsburgh, PA, 15261, USA.
³ Department of Pharmaceutical Sciences and Computational Chemical Genomics Screening Center, Pharmacometrics & Systems Pharmacology (PSP) Pharmacoanalytics, School of Pharmacy, University of Pittsburgh, 6411 Salk Hall, 3501 Terrace Street, Pittsburgh, PA, 15261, USA. ZHF11@pitt.edu.
⁴ National Center of Excellence for Computational Drug Abuse Research, University of Pittsburgh, Pittsburgh, PA, 15261, USA. ZHF11@pitt.edu.
⁵ Department of Pharmaceutical Sciences and Computational Chemical Genomics Screening Center, Pharmacometrics & Systems Pharmacology (PSP) Pharmacoanalytics, School of Pharmacy, University of Pittsburgh, 6411 Salk Hall, 3501 Terrace Street, Pittsburgh, PA, 15261, USA. yix49@pitt.edu.
⁶ National Center of Excellence for Computational Drug Abuse Research, University of Pittsburgh, Pittsburgh, PA, 15261, USA. yix49@pitt.edu.
⁷ Department of Pharmaceutical Sciences and Computational Chemical Genomics Screening Center, Pharmacometrics & Systems Pharmacology (PSP) Pharmacoanalytics, School of Pharmacy, University of Pittsburgh, 6411 Salk Hall, 3501 Terrace Street, Pittsburgh, PA, 15261, USA. xix15@pitt.edu.
⁸ National Center of Excellence for Computational Drug Abuse Research, University of Pittsburgh, Pittsburgh, PA, 15261, USA. xix15@pitt.edu.
⁹ Drug Discovery Institute and Departments of Computational Biology and Structural Biology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, 15261, USA. xix15@pitt.edu.

PMID: 40140128
DOI: 10.1007/s11095-025-03848-w

GraphDeep-hERG: Graph Neural Network PharmacoAnalytics for Assessing hERG-Related Cardiotoxicity

Yankang Jing et al. Pharm Res. 2025 Apr.

. 2025 Apr;42(4):579-591.

doi: 10.1007/s11095-025-03848-w. Epub 2025 Mar 26.

Authors

Affiliations

¹ Department of Pharmaceutical Sciences and Computational Chemical Genomics Screening Center, Pharmacometrics & Systems Pharmacology (PSP) Pharmacoanalytics, School of Pharmacy, University of Pittsburgh, 6411 Salk Hall, 3501 Terrace Street, Pittsburgh, PA, 15261, USA.
² National Center of Excellence for Computational Drug Abuse Research, University of Pittsburgh, Pittsburgh, PA, 15261, USA.
³ Department of Pharmaceutical Sciences and Computational Chemical Genomics Screening Center, Pharmacometrics & Systems Pharmacology (PSP) Pharmacoanalytics, School of Pharmacy, University of Pittsburgh, 6411 Salk Hall, 3501 Terrace Street, Pittsburgh, PA, 15261, USA. ZHF11@pitt.edu.
⁴ National Center of Excellence for Computational Drug Abuse Research, University of Pittsburgh, Pittsburgh, PA, 15261, USA. ZHF11@pitt.edu.
⁵ Department of Pharmaceutical Sciences and Computational Chemical Genomics Screening Center, Pharmacometrics & Systems Pharmacology (PSP) Pharmacoanalytics, School of Pharmacy, University of Pittsburgh, 6411 Salk Hall, 3501 Terrace Street, Pittsburgh, PA, 15261, USA. yix49@pitt.edu.
⁶ National Center of Excellence for Computational Drug Abuse Research, University of Pittsburgh, Pittsburgh, PA, 15261, USA. yix49@pitt.edu.
⁷ Department of Pharmaceutical Sciences and Computational Chemical Genomics Screening Center, Pharmacometrics & Systems Pharmacology (PSP) Pharmacoanalytics, School of Pharmacy, University of Pittsburgh, 6411 Salk Hall, 3501 Terrace Street, Pittsburgh, PA, 15261, USA. xix15@pitt.edu.
⁸ National Center of Excellence for Computational Drug Abuse Research, University of Pittsburgh, Pittsburgh, PA, 15261, USA. xix15@pitt.edu.
⁹ Drug Discovery Institute and Departments of Computational Biology and Structural Biology, School of Medicine, University of Pittsburgh, Pittsburgh, PA, 15261, USA. xix15@pitt.edu.

PMID: 40140128
DOI: 10.1007/s11095-025-03848-w

Abstract

Purpose: The human Ether-a-go-go Related-Gene (hERG) encodes rectifying potassium channels that play a significant role during action potential repolarization of cardiomyocytes. Blockade of the hERG channel by off-target drugs can lead to long QT syndrome, significantly increasing the risk of proarrhythmic cardiotoxicity. Traditional hERG screening methods are effort-demanding and time-consuming. Thus, it is essential to develop computational methods to utilize the existing knowledge for faster and more accurate in silico screening. Although with wide use of deep learning/machine learning algorithms, existing computational models often rely on manually defined atomic features to represent atom nodes, which may overlook critical underlying information. Thus, we want to provide a new method to learn the atom representation automatically.

Methods: We first developed an automated atom embedding model using deep neural networks (DNNs), trained with 118,312 compounds collected from the ZINC database. We then trained a Graph neural networks (GNNs) model with 7909 ChEMBL compounds as the classifying part. The integration of our atom embedding model and GNN models formed a classifier that could effectively distinguish between hERG inhibitors and non-inhibitors.

Results: Our atom embedding model achieved 0.93 accuracy in representing structures. Our best GNN model achieved an accuracy of 0.84 and outcompeted traditional machine-learning models, as well as published AI-driven models, in external testing.

Conclusions: These results highlight the potential of our automated atom embedding model as a standard for generating robust molecular representations. Its integration with advanced GNN algorithms offers promising assistance for screening hERG inhibitors and accelerating drug discovery and repurposing.

Keywords: HERG; alzheimer’s disease; atom-type embedding model; cardiovascular toxicity; deep learning/machine learning; graph neural networks.

PubMed Disclaimer

Conflict of interest statement

Declarations. Conflict of Interest: All authors have no conflict of interest to declare.

References

1. Gintant G, Sager PT, Stockbridge N. Evolution of strategies to improve preclinical cardiac safety testing. Nat Rev Drug Discov. 2016;15(7):457–71. - PubMed - DOI
1. Jing YK, Easter A, Peters D, Kim N, Enyedy IJ. In silico prediction of hERG inhibition. Future Med Chem. 2015;7(5):571–86. - PubMed - DOI
1. Trudeau MC, Warmke JW, Ganetzky B, Robertson GA. HERG, a Human Inward Rectifier in the Voltage-Gated Potassium Channel Family. Science. 1995;269(5220):92–5. - PubMed - DOI
1. Aronov AM. Predictive in silico modeling for hERG channel blockers. Drug Discov Today. 2005;10(2):149–55. - PubMed - DOI
1. Stockbridge N, Morganroth J, Shah RR, Garnett C. Dealing with Global Safety Issues Was the Response to QT-Liability of Non-Cardiac Drugs Well Coordinated? Drug Saf. 2013;36(3):167–82. - PubMed - DOI

MeSH terms

Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions

Substances

Actions
Actions
Actions
Actions

LinkOut - more resources

Full Text Sources
- Springer

Save citation to file

Email citation

Add to Collections

Add to My Bibliography

Your saved search

Create a file for external citation management software

Your RSS Feed

GraphDeep-hERG: Graph Neural Network PharmacoAnalytics for Assessing hERG-Related Cardiotoxicity

Affiliations

GraphDeep-hERG: Graph Neural Network PharmacoAnalytics for Assessing hERG-Related Cardiotoxicity

Authors

Affiliations

Abstract

Conflict of interest statement

References

MeSH terms

Substances

LinkOut - more resources

Full Text Sources