Biocomputing: Theory and Applications

Biological information processing is mediated by molecular interactions in terms of biochemical reactions. Superficially, the processes conjure up an imagery of random diffusion and collisions, at least in the solution phase of cytoplasm. Yet, human beings are capable of highly intelligent decision-making that exhibits no trace of randomness. Furthermore, advocates of classical physical determinism claim that biological determinism is absolute. This talk attempts to reconcile these two disparate pictures. The main issue is the control laws that govern the input-output relationship of biocomputing at three hierarchical levels, as stipulated by Michael Conrad's macroscopic-microscopic nested multi-level scheme of information processing, and an additional level of mescopic scheme at the membrane level.

Interactions of macromolecules in the cytoplasm appear more random when they are separated by large distances. In mesoscopic ranges, non-covalent bond interactions render the process more deterministic but never absolutely deterministic. For lack of better terminology, we call that relatively deterministic. The process is tantamount to heuristic searching in artificial intelligence. At the mesoscopic level, the dynamics of ion channel activation show that the control laws at the mesoscopic level are not even well-defined but are defined only probabilistically. Thus, endogenous noise is a crucial component for biocomputing, rather than unavoidable but tolerated interference. At the macroscopic level, however, the control law re-converges to a well-defined control law, as described by the Hodgkin-Huxley theory of action potential and the experimental data they generated. This level of computing is the basis of neural network computing. Nature's exploitation of endogenous noise casts a serious doubt on the validity of absolute biological determimsm. Nevertheless, a relative degree of determinism is preserved because of the re-convergence of the control laws at higher levels.

Ultimately, whether biological processes is deterministic or not is tied to the question whether physical determinism is absolute or not. A simple argument, based on elementary probablity concepts, is proposed to show that, contrary to common belief, absolute physical determinism (as well as microscopic reversibility) is contradictory to macroscopic irreversibility. If, however, microscopic reversiblity is assumed to be approximately true, macroscopic irreversibility can be understood in terms of chaos theory. Also to be shown is that Laplace's absolute physical determinism can neither be proved or disproved. Applications to the free will problem and human creativity will be discussed.