doi:10.1038/nindia.2012.10 Published online 23 January 2012

Virtual cell: Making a virtual heart

Can a human heart be reliably built upon thousands of mathematical equations? Will a human heart beating on a computer help in treating cardiac conditions? Pawan Dhar seeks the answers.

Model of a virtual heart.
© Pawan Dhar

The human heart is an amazing piece of biological machinery. Every second it choreographs a symphony of electrical signals, blood flow and muscular movements resulting in what we know as a heartbeat. The fist-size human heart beats 100,000 times a day on an average to send 7,500 litres of oxygen rich blood into a 96,000-km blood vessel network spread across the body.

Understanding heart physiology has enormous practical applications for human health. Damage to the heart could be life threatening. Fatal heart conditions range from coronary artery disease, angina and heart failure to arrhythmias. A large number of compounds used for treating heart conditions can cause fatal disruptions in the normal heart beat.

Mathematical models

To understand heart dynamics at the molecular level and treat pathogenic conditions, increasingly sophisticated mathematical models have been built over the last four decades. Back in 1952, the stage was set with Alan Hodgkin and Andrew Huxley's pioneering work that offered a mathematical view of excitable cells1. Their work of developing equations for nerve conduction in giant squid cells was so significant that it won them a Nobel Prize for physiology and medicine in 1963. Their paper correctly predicted the shape of action potential, impedance changes and the conduction velocity.

Encouraged by the paper, the 1960s decade saw a number of publications on the theme. The earliest was the discovery of calcium current in the heart2 followed by a detailed description of potassium channel dynamics3 and the development of Purkinje fibre model that eventually evolved into the first ventricular model cell4. Another important discovery in late 1960s was the finding of sodium-calcium exchange in cardiac muscle cells (Reuter and Seitz, 1969).

Dennis Noble's work

Encouraged by these fundamental discoveries, Dennis Noble, a British biologist at Oxford, pioneered the integration of data from various cells to build a virtual heart. The first three dimensional model of the heart was built on a University of Minnesota supercomputer in collaboration with Raimond Winslow. This virtual heart enabled a better understanding of ectopic heart beats5 and interactions between sinus node and atrial cells6.

Noble's goal was to model how heart worked as an integrated organ and what molecular processes evolved into arrhythmias i.e. irregular heartbeats.

After 1995, Noble's team was responsible for releasing several versions of virtual heart, integrating experimental data from various sources and going back and forth between experiment and computation. The virtual heart models are getting increasingly sophisticated with the inclusion mechanisms that maintain a constant pH in the cells, thereby ensuring an optimum metabolic function of cells.

As a part of Noble's collaborative network, several interesting developments7, 8, 9  in virtual heart community have taken place. A recent mechanical model that considers 3D geometry of heart tissues has been developed by an Auckland team led by Peter Hunter. Using this model, one can study how current flows around the heart. To make it more comprehensive and understand how nervous system impacts heartbeats, David Paterson's team at Oxford has modelled a neural network of the heart and added to the previous versions.

One can now ask questions such as: how does a genetic mutation lead to a change in electrophysiology of a cardiac cell.

Pharma connection

The pharma industry is extremely interested in the virtual heart project. Several initiatives to study arrhythmias and ischemic heart conditions are going on. The rise of sodium concentration by the persistent sodium current in ischaemia has been correctly predicted by computational models.

However, given that approximately 40% of the drugs produce cardiac arrhythmias, heart failure is one of the biggest reasons of drugs failing in clinical trials. This is due to one of the membrane proteins which interacts with the drugs. Due to this reason, in 2004 Merck had to withdraw its key product — an arthritis painkiller Vioxx and suffer a massive financial loss. Vioxx induced heart attack in some people.

Failures & limitations

Currently, computer models of biologic processes do not behave exactly like the real heart. An iterative interaction between experiment and simulation is therefore important to find the best modelling scenarios and boundary conditions. Due to vast improvement in technology, the calculations that required several hours of time in the 1960s can be run in microseconds on a small laptop.

However, it is not yet possible to capture "all the channels in all cardiac cells in all the environmental contexts". This is too big a computational problem. Therefore, one must look for appropriate questions that can be answered by simulating a virtual heart at a certain level of abstraction. In many cases, arrhythmias that look similar in their profile can have different causal routes.

Future work

Heart models have become increasingly sophisticated over the last four decades. A Wellcome-funded consortium is looking at taking the virtual heart as close to the real heart. The key is to collect and integrate ion channel data of as many cell types as possible. It is interesting that ion channel dynamics modelled in one tissue may be applicable to another tissue.

However, the big systems biology question still remains — will human physiology become a computational problem in future? Time will tell.

This article is the fourth in a series entitled 'Virtual Cell'.


  1. Hodgkin, A. L. et al. A quantitative description of membrane current and its application to conduction and excitation in nerve. J. Physiol. 117, 500-544 (1952) | PubMed | ISI |
  2. Reuter, H. The dependence of slow inward current in Purkinje fibres on the extracellular calcium concentration. J. Physiol. 192, 479-492 (1967) | PubMed | ISI |
  3. Noble, D. et al. The kinetics and rectifier properties of the slow potassium current in cardiac Purkinje fibres. J. Physiol. 195, 185-214 (1968) | PubMed |
  4. McAllister, R. E. et al. Reconstruction of the electrical activity of cardiac Purkinje fibres. J. Physiol. 251, 1-59 (1975) | PubMed |
  5. Winslow, R. et al. Generation and propagation of triggered activity induced by spatially localised Na-K pump inhibition in atrial network models. Proc. Biol. Sci. 254, 55-61 (1993)  | Article | PubMed |
  6. Noble, D. et al. Propagation of pacemaker activity: interaction between pacemaker cells and atrial tissue. Pacemaker Activity and Intercellular Communication, ed. Huizinga, J. D. 73-92 (1995) CRC Press, Boca Raton, FL, USA
  7. Noble, D. Modelling the heart: insights, failures and progress. Bioessays 24, 1155-1163 (2002) | Article | PubMed | ISI |
  8. Mirams, G. R. et al. Is it time for in silico simulation of drug cardiac side effects? Ann. N. Y. Acad. Sci. 1245, 44-47 (2011) | Article | PubMed |
  9. Noble, D. Successes and failures in modeling heart cell electrophysiology. Heart Rhythm 8, 1798-1803 (2011) | Article | PubMed | ISI |