Mathematical modeling in systems toxicology enables a comprehensive understanding of the effects of pharmaceutical substances on cardiac health. However, the complexity of these models limits their widespread application in early drug discovery. In this article, we introduce a novel approach to solving parameterized models of cardiac action potentials (CAPs) by combining meta-learning techniques with systems biology-informed neural networks (SBINNs). The proposed method, hyperSBINN, effectively addresses the challenge of predicting the effects of various compounds at different concentrations on CAPs, outperforming traditional differential equation solvers in speed. Our model efficiently handles scenarios with limited data and complex parameterized differential equations. The hyperSBINN model demonstrates robust performance in predicting APD90 values, indicating its potential as a reliable tool for modeling cardiac electrophysiology and aiding in preclinical drug development. This framework represents an advancement in computational modeling, offering a scalable and efficient solution for simulating and understanding complex biological systems.
AlberM, Buganza TepoleA, CannonWR, et al.Integrating machine learning and multiscale modeling—perspectives, challenges, and opportunities in the biological, biomedical, and behavioral sciences. NPJ Digit Med, 2019; 2(1):115.
2.
AlievRR, PanfilovAV. A simple two-variable model of cardiac excitation. Chaos Solitons Fractals, 1996; 7(3):293–301.
3.
AndrásV, TomekJ, NagyN, et al.Cardiac transmembrane ion channels and action potentials: Cellular physiology and arrhythmogenic behavior. Physiol Rev, 2021.
4.
BischofR, KrausMA. Multi-objective loss balancing for physics-informed deep learning. Comput Methods Appl Mech Eng, 2025; 439:117914.
5.
CarmelietE, MubagwaK. Antiarrhythmic drugs and cardiac ion channels: Mechanisms of action. Prog Biophys Mol Biol, 1998; 70(1):1–72.
6.
CentofantiE, GhiottoM, PavarinoLF. Learning the Hodgkin–Huxley model with operator learning techniques. Comput Methods Appl Mech Eng, 2024; 432:117381.
7.
ChauhanVK, ZhouJ, LuP, et al.A brief review of hypernetworks in deep learning. Artif Intell Rev, 2024; 57(9):250.
8.
ChiuC-E, PintoAL, ChowdhuryRA, et al.Characterisation of anti-arrhythmic drug effects on cardiac electrophysiology using physics-informed neural networks. arXiv Preprint, 2024.
9.
CorriasA, GilesW, RodriguezB. Ionic mechanisms of electrophysiological properties and repolarization abnormalities in rabbit Purkinje fibers. Am J Physiol Heart Circ Physiol, 2011; 300(5):H1806–H1813.
10.
CuomoS, Di ColaVS, GiampaoloF, et al.Scientific machine learning through physics-informed neural networks: Where we are and what’s next. J Sci Comput, 2022; 92(3):1–27.
11.
DanekerM., ZhangZ., KarniadakisG. E., and LuL. Systems biology: Identifiability analysis and parameter identification via systems-biology-informed neural networks. In: Computational Modeling of Signaling Networks. Springer; 2023; pp. 87–105.
12.
de Avila Belbute-PeresF, ChenY-F, ShaF. HyperPINN: Learning parameterized differential equations with physics-informed hypernetworks. In: The Symbiosis of Deep Learning and Differential Equations. NeurIPS; 2021.
13.
GangloffH, JouvinN. Jinns: A JAX library for physics-informed neural networks. arXiv Preprint, 2024.
14.
GaoB, YaoR, LiY. Physics-informed neural networks with adaptive loss weighting algorithm for solving partial differential equations. Comput Mathematics Applications, 2025; 181:216–227; doi: 10.1016/j.camwa.2025.01.007
15.
GerlachU. channel blockers: Potential antiarrhythmic agents. Drugs of the Future, 2001; 26(5):473–484.
16.
GranditsT, PezzutoS, CostabalFS, et al.Learning atrial fiber orientations and conductivity tensors from intracardiac maps using physics-informed neural networks. In: International Conference on Functional Imaging and Modeling of the Heart. Springer; 2021; pp. 650–658.
Herrero MartinC, OvedA, ChowdhuryRA, et al.EP-PINNS: Cardiac electrophysiology characterisation using physics-informed neural networks. Front Cardiovasc Med, 2021; 8:768419.
19.
HuangDZ, NelsenNH, TrautnerM. An operator learning perspective on parameter-to-observable maps. arXiv Preprint, 2024.
20.
JenkinsonS, SchmidtF, RibeiroLR, et al.A practical guide to secondary pharmacology in drug discovery. J Pharmacol Toxicol Methods, 2020; 105:106869.
21.
JiW, QiuW, ShiZ, et al.Stiff-PINN: Physics-informed neural network for stiff chemical kinetics. J Phys Chem A, 2021; 125(36):8098–8106.
22.
KingmaDP, BaJ. Adam: A method for stochastic optimization. arXiv Preprint, 2014.
23.
KrishnapriyanA, GholamiA, ZheS, et al.Characterizing possible failure modes in physics-informed neural networks. arXiv Preprint, 2021.
24.
LeeH-M, YuM-S, KazmiSR, et al.Computational determination of hERG-related cardiotoxicity of drug candidates. BMC Bioinformatics, 2019; 20(Suppl 10):250–273.
25.
LiL, TaiX-C, ChanRH-F. A new method to compute the blood flow equations using the physics-informed neural operator. J Comput Phys, 2024; 519:113380.
MayourianJ, SobieEA, CostaKD. An introduction to computational modeling of cardiac electrophysiology and arrhythmogenicity. In: Experimental Models of Cardiovascular Diseases: Methods and Protocols. Springer; 2018; pp. 17–35.
28.
MohrM, ChambardJ-M, BalletV, et al.Accurate in silico simulation of the rabbit Purkinje fiber electrophysiological assay to facilitate early pharmaceutical cardiosafety assessment: Dream or reality? J Pharmacol Toxicol Methods, 2022; 115:107172.
PellegriniL, GhiottoM, CentofantiE, et al.Learning high-dimensional ionic model dynamics using Fourier neural operators. arXiv Preprint, 2025.
31.
PrzedborskiM, SmalleyM, ThiyagarajanS, et al.Systems biology informed neural networks (sbinn) predict response and novel combinations for pd-1 checkpoint blockade. Commun Biol, 2021; 4(1):877.
32.
RadhakrishnanK. and HindmarshA. C. Description and use of lsode, the livermore solver for ordinary differential equations. Technical report, National Aeronautics and Space Administration; 1993.
33.
RaissiM, PerdikarisP, KarniadakisGE. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. J Comput Phys, 2019; 378:686–707.
34.
Sahli CostabalF, YangY, PerdikarisP, et al.Physics-informed neural networks for cardiac activation mapping. Front Phys, 2020; 8:42.
35.
TrayanovaNA, LyonA, ShadeJ, et al.Computational modeling of cardiac electrophysiology and arrhythmogenesis: Toward clinical translation. Physiol Rev, 2024; 104(3):1265–1333.
36.
WilliamsG, MiramsGR. A web portal for in-silico action potential predictions. J Pharmacol Toxicol Methods, 2015; 75:10–16.
37.
WuC, ZhuM, TanQ, et al.A comprehensive study of non-adaptive and residual-based adaptive sampling for physics-informed neural networks. Comput Methods Appl Mech Eng, 2023; 403:115671.
38.
YazdaniA, LuL, RaissiM, et al.Systems biology informed deep learning for inferring parameters and hidden dynamics. PLoS Comput Biol, 2020; 16(11):e1007575.
