Low power and massively parallel simulation of oscillatory biochemical networks on FPGA

Abstract

Biological functions emerge from a multitude of chemical species woven into intricate biochemical networks. It is crucial to compute the dynamics of a biochemical network from its kinetics and topology. In order to reverse engineer networks and map their design space, dynamics needs to be simulated for many different parameters and topologies, leading to a combinatorial explosion that requires heavy computational power. To solve this issue, we show here an application of FPGA platform to simulate biochemical networks. As a toy model, we simulate a structurally simple network with a rich oscillatory dynamics: a predator-prey biochemical oscillators. The network mimics predator-prey dynamics. We show that FPGA can simulate the dynamics of PP faithfully. These results open the door to more energy-efficient simulations.