Many mathematical problems, e.g., cryptography, network routing, require the exploration a large number of candidate solutions. Because the time required for solving these problems grows exponentially with their size, electronic computers, which operate sequentially, cannot solve them in a reasonable time. In contrast, biological organisms routinely process information in parallel for essential tasks, e.g., foraging, searching for space opens three possible biocomputing avenues.
Biomimetic algorithms translate biological procedures, e.g., space searching, chemotaxis, etc., into mathematical algorithms. This approach was used to derive fungi-inspired algorithms for searching space and bacterial chemotaxis-inspired algorithms for finding the edges of geometrical patterns.
Biosimulation uses the procedures of large numbers of motile biological agents, directly, without any translation to formal mathematical algorithms, thus by-passing computation-proper. The agents explore complex networks that mimic real situations, e.g., traffic. This approach focused almost entirely on traffic optimization, using an amoeboid organisms placed in confined geometries, with chemotactic ‘cues’, e.g., nutrients in node coordinates.
Computing with biological agents in networks uses very large number of agents exploring microfluidics networks, purposefully designed to encode hard mathematical problems. The foundations of a parallel-computation system in which a combinatorial problem (SUBSET SUM) is encoded into a graphical, modular network embedded in a nanofabricated planar device was reported. Exploring the network in a parallel fashion using a large number of independent agents, e.g., molecular motor-propelled cytoskeleton filaments, bacteria, algae, solves the mathematical problem. This device uses orders of magnitude less energy than conventional computers, additionally addressing issues related to parallel computing implementation.