Complex living systems would benefit from being modelled as if they were computer programs.
Biologists have amassed huge data sets, such as the entire human genome, but how the components work together is often a mystery. Most simulations require detailed knowledge of the rates at which components change, which often aren't known.
Now Jasmin Fisher of Microsoft Research in Cambridge, UK, and Thomas Henzinger of the Federal Polytechnic School of Lausanne, Switzerland, suggest using models structured like computer programs, where many "subroutines" run in parallel and produce outputs that depend on each other . By representing proteins, say, as subroutines, it is possible to work out how the overall system works without knowing some details.