dorsal/arxiv
View SchemaEnsembles of Protein Molecules as Statistical Analog Computers
| Authors | Victor Eliashberg |
|---|---|
| Categories | |
| ArXiv ID | physics/0308041 |
| URL | https://arxiv.org/abs/physics/0308041 |
Abstract
A class of analog computers built from large numbers of microscopic probabilistic machines is discussed. It is postulated that such computers are implemented in biological systems as ensembles of protein molecules. The formalism is based on an abstract computational model referred to as Protein Molecule Machine (PMM). A PMM is a continuous-time first-order Markov system with real input and output vectors, a finite set of discrete states, and the input-dependent conditional probability densities of state transitions. The output of a PMM is a function of its input and state. The components of input vector, called generalized potentials, can be interpreted as membrane potential, and concentrations of neurotransmitters. The components of output vector, called generalized currents, can represent ion currents, and the flows of second messengers. An Ensemble of PMMs (EPMM) is a set of independent identical PMMs with the same input vector, and the output vector equal to the sum of output vectors of individual PMMs. The paper suggests that biological neurons have much more sophisticated computational resources than the presently popular models of artificial neurons.
{
"annotation_id": "1d463797-f7e5-488d-90be-1dca14bbe980",
"date_created": "2026-03-02T18:00:46.503000Z",
"date_modified": "2026-03-02T18:00:46.503000Z",
"file_hash": "0f776e1906f415485b351d0a6ef8c2ce4a0d2de608aec5ee66ea6724ab4f24ac",
"private": false,
"record": {
"abstract": "A class of analog computers built from large numbers of microscopic\nprobabilistic machines is discussed. It is postulated that such computers are\nimplemented in biological systems as ensembles of protein molecules. The\nformalism is based on an abstract computational model referred to as Protein\nMolecule Machine (PMM). A PMM is a continuous-time first-order Markov system\nwith real input and output vectors, a finite set of discrete states, and the\ninput-dependent conditional probability densities of state transitions. The\noutput of a PMM is a function of its input and state. The components of input\nvector, called generalized potentials, can be interpreted as membrane\npotential, and concentrations of neurotransmitters. The components of output\nvector, called generalized currents, can represent ion currents, and the flows\nof second messengers. An Ensemble of PMMs (EPMM) is a set of independent\nidentical PMMs with the same input vector, and the output vector equal to the\nsum of output vectors of individual PMMs. The paper suggests that biological\nneurons have much more sophisticated computational resources than the presently\npopular models of artificial neurons.",
"arxiv_id": "physics/0308041",
"authors": [
"Victor Eliashberg"
],
"categories": [
"physics.bio-ph",
"cs.AI",
"cs.NE",
"physics.comp-ph",
"physics.data-an",
"q-bio.NC"
],
"title": "Ensembles of Protein Molecules as Statistical Analog Computers",
"url": "https://arxiv.org/abs/physics/0308041"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "4c0a5a69-07a8-4ae8-9990-85ee7d6197ee",
"id": "arXiv Dataset IDs",
"type": "Model",
"variant": "snapshot-2026-03-01",
"version": "0.1.0"
},
"user_id": 1000002
}