|
|
|
|
|
|
|
| Dateline: June 25, 2000 | |
|
|
|
|
|
|
It is true that neural network simulations have dramatically improved our computers' abilities for empirical classification and pattern recognition; however, as of now no classical neural network model has the algorithm used by real neurons for determining exactly when to fire (termed activation function), or more importantly exactly when and how much of the neurotransmitters are secreted into the synapse. It is entirely possible that complex computations are carried out within a single neuron, making the brain massively parallel. While on the microscale the axon clearly functions as a kind of wire that propagates an electric current to the synapse, on the nanoscale it serves the purpose of transporting vesicles containing the neurotransmitters required for signal transduction across the synapse.
The vesicles (nanoscale vessels or containers) are transported by the actomyosin molecular motors described on the previous page. Actin and myosin are the two components of actomyosin, which is a ubiquitous complex used for muscle contraction in addition to its vesicle transport function in neurons. One Adenosine Triphosphate (ATP) molecule is used as the molecular fuel for each step made by myosin along the actin filament, resulting in ~60% efficiency of 'chemomechanical power transduction' [5] in muscle cells. The release of energy from the ATP molecule occurs at an extremely slow rate (10-2s/ATP molecule) and is thought to proceed by the emission of a sequence of quanta [6].
The main arguement of Tegmark against the possibility of the brain acting as a quantum computer relied on his caluclation of the decoherence time of a kink in a microtubule being 10-13s. He claimed that since the neuron operates on a time scale between 10-3s and 10-1s, the decoherence is too fast [1]. This may be true for microtubule kinks, but the effective freezing of the myosin engine core operates on the appropriate time scale. Furthermore, Matsuno has determined the momentum of the condensed quantum state from the sliding velocity of the myosin along an actin filament, which yields a de Broglie wavelength of 4.5 nm. Since this length is sufficiently greater than the 2.5 nm diameter of each actin monomer it is likely that the quantum coherence and entanglement extends over several actin monomers aligned along the actin filament [4]. These facts combinded with the near absolute zero effective temperature suggests that the actomyosin vessicle transport system does exhibit the quantum coherence and entanglement necessary for quantum computation within a single neuron.
Next Page: Novel Biophysical Methods
References:
[5] Robert A. Freitas Jr. Nanomedicine Volume I: Basic Capabilities, Landes Bioscience 1999, p 147. Table of Contents
[6] Koichiro Matsuno, 'Being free from ceteris paribus a vehicle of founding physics upon biology rather than the other way around' Appl. Math. Comp. (1993) 56, 261-279. Koichiro Matsuno's Homepage
Keywords: quantum biology myosin coherence neural network simulation massively parallel nanoscale vessicle transport signal transduction actomyosin molecular motors actin ATP adensine triphosphate chemomechanical power transduction rate freitas matsuno neuron quantum computation