Electronic modeling of biochemical pathways on reconfigurable computing platform-a repressive circuit of soil bacteria

Cytomorphic hardware, in which bio-cellular processes are mapped onto an electronic substrate, has gained quite a prominence in the past few decades due to advancements in the fields of systems biology, synthetic biology, and other related disciplines. Conventionally, the analog substrate is the preferred choice for nonlinear biochemical reactions in the biological circuits; however, many researchers have exploited the flexibility, reconfigurability, and ease of design of digital platforms such as reconfigurable field-programmable gate arrays (FPGAs). In this work, we briefly examined work on such reconfigurable platforms and proposed a novel and straightforward technique to implement the systems of ordinary differential equations describing such biochemical reactions in a modular fashion using single-precision floating-point intellectual property cores available on a representative reconfigurable platform. A simplified biochemical system of a repressive three-way network of soil bacteria was taken into account, and the system of ordinary differential equations was first simulated in a software-based environment with results compared against a hardware-based realization. We observed a significant speed-up of up to 5x in the reconfigurable platform-based realization compared to the software-based realization. The values in the hardware-based realization were also more accurate compared to the software-based approach. Statistical error metrics (RMSE, MAE, and NRMSE) further confirmed a close numerical match between the two implementations. Hence, the results conformed well, and thus, this design strategy can be incorporated into the future cytomorphic design on field-programmable gate arrays.