Breaking the Box: Simulated Protein Computing

Computers since the 1940s have shared the same basic architecture described by Turing and von Neumann, in which one central processor has access to one contiguous block of main memory. This architecture is challenged by modern applications that require greater parallelism, distribution, coordination, and complexity. Here we show that a model of protein interactions can serve as a new architecture, performing useful calculations in a way that provides for much greater scalability, flexibility, adaptation, and power than does the traditional von Neumann architecture. We found that even this simple simulation of protein interactions is universal, being able to replicate the calculation performed on a digital computer, yet without relying upon a central processor or main memory. We anticipate that the convergence of information- and life-sciences is poised to deliver a platform that invigorates computing as it provides insight into the complexity of living systems.