Can evolution be calculated?
Noro viruses are attacking, influenza is booming, bird flu is approaching again - wouldn't it be nice to be able to simulate such epidemics and evolutionary processes so that we know in good time what's in store for us? Darwin in the box, so to speak? It doesn't have to involve all species on earth right away. Even the evolution of all influenza viruses or even worse culprits might offer the chance to save lives through better prophylaxis.
When I was still in Oxford on my doctoral thesis on black holes, I occasionally chatted with one of the young geniuses there about theoretical biology at nearby Magdalen College. At that time, new models for population dynamics were just emerging, and biochemists were already trying to simulate biomolecules in liquids or the folding of proteins – a tedious, even impossible business on the Stone Age computers of the 1970s.
Later autocatalytic processes became popular: Manfred Eigen and Peter Schuster constructed and simulated their hypercycles. Nonlinearities became fashionable, the butterfly effect was born, thanks to a climatologist's rather innocuous nonlinear equation, which was solved iteratively.
Then cellular automata appeared, with which certain propagation and development processes could be represented, as also described by Manfred Eigen's "The Game": all sorts of crab- or star-shaped structures that crawl across the plain and devour each other. And finally, Stephen Wolfram suggested in "A new kind of science" (2002) that the world is actually just a kind of cellular automaton. Bioinformaticians now dominate the field, they check molecules on the computer, they test their vehicles in crash tests like car manufacturers. No wonder. Anyone who can shorten the path to a new drug with such a design of drugs or vaccines no longer needs to worry. And protein folding is still unsolved, despite all the supercomputers and teraflops.
But what about evolution? Could the next bird flu epidemic not only be feared, but also be better predicted thanks to realistic simulation? And what can physicists contribute to this? Michael W. Deem has taken on the question (Physics Today 1/2007, p. 42) and thinks that mathematics can do a lot, whether it's biological evolution or pathogen development. However, biological systems have their peculiarities that are unusual in physics or chemistry.
Difficulty number one: Cells with their inhomogeneously distributed biomolecules are subject to particularly large fluctuations. Also, some of these molecules, such as DNA, mRNA or enzymes, usually occur in such small numbers in the cell that the usual equations of solution chemistry are not applicable. Transport and reaction of individual molecules dominate the action.
Difficulty number two: Biological systems are rarely in thermodynamic equilibrium. Many reactions only occur at certain times, far from equilibrium and in small populations - room for new modeling, for example with the help of statistical physics.
Key Application: How do resistant pathogens develop, whether in influenza or tuberculosis? How can viruses spread from animals to humans? According to American he alth researchers, 12 of the last 13 viral diseases observed in humans spread from animals to humans. Especially in the case of Sars and the H5N1 bird flu virus, it would be extremely important to understand the details of such adaptation processes.
There is no doubt that theoretical physicists who are passionate about the cosmos or the interior of elementary particles are beginning to apply their tools more to issues that directly affect humans.
Reinhard Breuer
