Virtual humans: a quest

In 1741 the automata-maker Jacques Vaucanson gave a lecture at Lyon’s Academy of Art.

The minutes of the meeting record that Vaucanson told of an amazing new project he had imagined ‘that of constructing an automaton figure which will imitate in its movements animal functions, the circulation of the blood, respiration, digestion, the combination of muscles, tendons nerves, etc.’

This automaton would be designed not to entertain but to enlighten: through experiments upon this artificial man physicians would gain new insights into human health, disorders and illness.

Whilst Vaucanson would struggle for 20 years to make his automaton, ultimately he would have to admit defeat. He was a man ahead of his time: the materials and tools required to build such a creation still elude us.

Yet that isn't the end of the story as one tool, the computer, promises one day to make Vaucanson's vision a (virtual) reality.

Oxford is at the forefront of efforts to create the Virtual Physiological Human (VPH). In essence this would be Vaucanson's automata on computer ('in sillico'), giving researchers from around the world the chance to share and combine their ideas and observations about how the human body works.

The challenges are immense. We still don't understand many of the fundamental physical processes that determine how our organs function and interact, especially at a cellular level. Despite this, recent advances in computing power and computational modelling mean scientists are starting to create models that quantify such processes.

I'll blog more about all the projects involved in due course, and the collaborations between Oxford's Computing Laboratory and Department of Physiology Anatomy and Genetics, but I thought I'd get straight to the heart of the matter:

preDICT is one of the projects just starting that has benefited from some forward-looking European Commission funding. Its stated mission is to model, simulate, and ultimately predict the impact of pharmacological compounds on the heart's rhythm using computer models.

To do this the international team, involving scientists from Oxford, will have to create mathematical models of the ion channels that control how and when heart cells contract, tissue models which capture the complex physical and chemical interactions; all wrapped up in computer code that can run such a model faster than a heart beating in realtime.

It's a big ask - preDICT's project manager Katherine Fletcher tells me they will still need to build a computer at the end of the project that's powerful enough to run the model! - yet there's no doubt that meaningful testing of drugs 'in sillico' would be of immense benefit to those developing new treatments.