Species Minimization in Computation with Biochemical Reactions

Engineering biochemical reactions for computational purposes is a common pursue in synthetic biology [1]. In such design tasks, molecular species have to be carefully engineered to ensure modularity and orthogonality [3, 7], and are scarce resources. Minimizing the number of involved molecular species is crucial to accomplish a complex computation within a confined biochemical environment. This report aims at species minimization by reusing modular and regular reactions in an asynchronous time-multiplexed fashion. Our method enhances not only species utility, but also re-programmability and robustness in realizing various logic circuits. A case study demonstrates the ease of design in realizing general logic computation, and simulation confirms the feasibility and robustness of the proposed method. The quest for deciphering nature’s design principle of living organisms has been the primary subject in systems biology. This goal cannot be fully achieved unless human engineers can construct complex biological systems from scratch. A synthetic approach to biology aims to create complex systems bottom up from elementary components [1]. In such construction, a convenient way of modeling biological systems is from a computational aspect regarding system states in terms of molecular concentrations. Despite the intrinsic hybrid nature (involving both continuous and discrete state evolutions) of a biological system, the dominating digital design methodology of electronic circuits can be systematically carried to designing computation with biochemical reactions [2, 5]. Building upon our prior work [2], we address a fundamental limitation of biological circuit design. Unlike electronic circuit design, where a logic gate can be freely instantiated into several copies to compose a complex logic function, in a biological circuit every signal corresponds to a distinct molecular species, and every instantiation of a logic gate “consumes” molecular species. As molecular species have to be carefully engineered to ensure modularity and orthogonality for proper operation [3, 7], they are expensive resources. It raises an important question, how to minimize molecular species involved in achieving a target computational task.