A new computational model developed by MIT researcher Jean-Jacques Slotine and colleagues allows better control of complex networks, such as the genes that regulate cellular metabolism, by identifying critical points that can be used to control the entire system.

The number of points that need to be controlled to influence the system depends on whether the network is densely or sparsely connected, they report in the May 12 cover story of Nature.

The model predicts that controlling sparse networks like gene regulatory networks requires control more than 80% of all points in the network, while controlling dense networks like neural networks in the brain requires only 10%.

This latest improvement on the half-century old field of control theory will have diverse applications for controlling any network through carefully targeted inputs.

Yang-Yu Lu et al., Controllability of complex networks, Nature 473, 167–173 (12 May 2011) doi:10.1038/nature10011