The phrase "virtual human" conjures up images of online avatars, or the pixelated people and polygon personas created for the next Hollywood blockbuster. But in biology, it refers to something rather more exciting — the way computers enable us to build lifelike models of organs, down to the last cell, that are providing extraordinary new insights into our bodies and minds.
As long ago as 1960, pioneering work by Professor Denis Noble, of Oxford University, gave us the first mathematical models of how cardiac cells behaved. Within two decades, the complete range of electrical, chemical, and mechanical processes behind the contraction of a heart muscle cell had been captured in the form of around 30 equations.
Using them, it became possible to create a computer model of a beating heart cell and then, eventually, to simulate the pulse of the billions of cells in a virtual heart. For more than a decade, this artificial heart has been used to screen promising cardiac medicines.
And recently, a computer model of blood flow overturned the textbook wisdom about sickle-cell disease, a serious hereditary blood disorder that affects more than 10,000 people in Britain and 100,000 in the US.
Traditional accounts of the disease hold that, while normal red blood cells are disc-shaped, enabling them to pass easily through blood vessels, the crescent-shaped blood cells produced in those who have the disease clog the capillaries, causing searing pain and tissue damage.
Now, however, Prof George Karniadakis at Brown University in Rhode Island has come up with a model of cell characteristics and how they affect blood flow. It shows that the rigid, crescent-shaped red blood cells don't cause these blockages on their own. It turns out that the disease produces four different types of red blood cells.
The model shows that one of these — SS2, which is softer and deformable — triggers the problem by sticking to capillary walls. Only then do the rigid sickle-shaped cells block the vessels themselves. The research, published last month, could provide a way to test drug treatments aimed at easing or preventing sickle cell.
"If a drug is designed to target the cells' adhesive properties, or if it's trying to make cells more flexible, we can put it to the test," explains Prof Karniadakis. The next great challenge for the mathematical modellers is to come up with a working version of the brain, with its billions of interconnected neurons.
Last week, a model of the world's smallest brain — that of a 1mm-long nematode worm, C. elegans, containing just 302 neurons — went on show last week at MOSI (the Museum of Science & Industry in Manchester).
Meanwhile, the Human Brain Project, led by Prof Henry Markram at the Ecole Polytechnique Federale de Lausanne on the shores of Lake Geneva, passed a key milestone last September, with a proof of concept that showed it could wire up virtual neurons in a way that matches what's been observed in studies of living nerve tissue.
The project will see its official launch this October, after the European Union decided to spend a billion euros to simulate the organ's 86 billion or so neurons. And even though the work will take more than a decade, the underlying research is already providing profound insights.
For instance, for almost a century scientists have been studying brain waves — yet how they originate from the activity of our interconnected neurons has been mysterious. But a few days ago, in the journal Neuron, the Markram team, working with colleagues at the Allen Institute for Brain Science in Seattle, reported that they had simulated the waves seen in a living rat brain using a detailed computer model of a neural circuit of 12,000 rat neurons.
"The electrical signals produced by the computer simulation and what was actually measured in the rat brain have some striking similarities," says Costas Anastassiou, of the Allen Institute. Another aspect of the project concerns blood flow in the brain, and the rest of the body, which will be handled by the Virtual Physiological Human, another European Union initiative.
This will run until 2020 - but using a simulation on HECToR, the UK's flagship supercomputer, based in Edinburgh, it is already possible to map the forces that flowing blood exerts on a weak vessel, and hence work out the risk of the ruptures that lead to an aneurysm.
The most exciting, and challenging, aspect of this work is that, in a few decades' time, as our understanding expands and computer power improves, it will become possible to simulate a full virtual body — and not just any body, but the body of a particular patient, from mitochondria to liver to nervous system.
When faced with risky surgery, or an unproven drug, doctors will be able to use it on this digital doppelganger first, before trying it out on the actual patient. This, in turn, leads to a further question: how far do we want to rely on our digital body double to find out what lies ahead? "There are novel legal and ethical issues raised by all of this," says Prof Peter Coveney of University College London, one of the leaders in this field.
"When virtual humans are common, we will all have to think about how much we want them to tell us." And, ultimately, we will be able to ask our avatar the biggest question of all: "How I am likely to die?"
Roger Highfield is the director of external affairs at the Science Museum.