EONS stands for Elementary Objects of Neural Systems. This program originated at The University of Southern California (Los Angeles) Brain Program around 1996. It appears oriented toward simulating medium-large neural networks (e.g., 800 neurons) using fairly detailed neuron and synaptic models (including a variety of specified ion channels involved in generating a psp). There are two major components to the EONS program: an EONS library and a user interface. The library contains neuron, synapse, synaptic cleft, postsynaptic spine, receptor channels, voltage-gated ion channels, and neural network. EONS has been used successfully to validate a hypothesis that receptor channel aggregation on the postsynaptic membrane is the cellular mechanism underlying the expression of long-term potential. Although not specifically stated in the "Summary Description," EONS probably is freeware and runs on UNIX systems. To learn more, visit the URL: http://www-hbp.usc.edu/Projects/eons.htm.

9.5.2 SNNAP

SNNAP is the acronym for Simulator for Neural Networks and Action Potentials. It was developed around 1994 at the University of Texas Health Science Center in Houston to do detailed, realistic modeling of single neurons and small neural networks. SNNAP runs under the UNIX environment; it was developed using ANSI C and Xlib. A DOS/Windows version exists that can simulate up to 20 neurons. Both versions have GUIs. UNIX SNNAP allows the user to simulate the injection of external currents into multiple cells, to simulate ion channel blocking by drugs by removing specific conductances, to modulate membrane currents with modulatory transmitters, and to simulate the voltage clamping of cells. Some 16 sample simulations can be downloaded, including burst generation, CPGs, etc. This appears to be an easy-to-use simulation program, and may be suited for use in an introductory neurobiology class. For details, see the URL: http://snnap.med.uth.tmc.edu/ /overview.htm. Questions and inquiries about obtaining SNNAP can be e-mailed to: [email protected].

9.5.3 SONN

SONN stands for Simulator of Neural Networks. SONN was developed at the Hebrew University, Jerusalem, Israel, and, like most neural modeling software, runs under the UNIX operating system and its variants. A new PC/Windows version was released in August 1999. SONN v1.0 can be downloaded free; however, the UNIX SONN manual costs U.S.$40.

SONN is oriented toward the fairly detailed simulation of single neurons and small neural networks. It allows specification of the following functional units: soma, axon; presynapse, postsynapse. Interested readers should contact http://icnc.huji. ac.il/Research/neuro.html, and http://www.Is.ac.il/~litvak/Sonn/sonn.html. Email requests for the manual to: [email protected] or [email protected].

Nodus v3.2 is a state-of-the-art neural modeling program originating in Belgium that has been written to run on Apple Macintosh™ computers (specifically, a Power Mac running System 7, or a Mac II series SE30, Centris 650/660, any Quadra with S 4 MB RAM and OS 6 or 7). Three different versions of Nodus v3.2 are available to run on various versions of Mac (3.2, 3.2P, and 3.2Q). See the URL: http://bbf-www.uia.ac.be/SOFT/NODUS_system.html for details. To quote from the Nodus Web information blurb:

Nodus combines a powerful simulator with sophisticated model database management. Models are defined in separate files: conductance definition files, neuron definition files and network definition files. All files specifying one model are linked together in a hierarchical structure and [are] automatically loaded when the top file is opened. Several conductance and neuron files can be open at the same time. A simulation database is build [sic] from user specified definition files and can be saved in a separate file, together with specific settings for graphic or text output, experiments, etc.

Two integration methods are available, a fifth-order Runge-Kutta/Fehlberg and a fast-forward Euler integrator, both with variable time steps. The value of any simulation database parameter can be manipulated by the user during a simulation. Networks can be "hard-wired" with up to 200 neurons and a maximum of 60 synapses with delays and/or 20 electric connections for each neuron. Currents of various waveforms can be injected in any compartment of the model. Two neurons can be simultaneously voltage-clamped. Selected ionic currents can be blocked to emulate drug action (e.g., TTX and TEA). Up to 13 ionic conductances can be simulated. Synaptic neurotransmitter release can be constant, voltage-dependent, or concentration pool dependent.

Postsynaptic conductance changes can follow the standard alpha model, with the synaptic input an impulse function, g(t) = gmax(t/x1) exp(1 - t/x1)], or follow a time course determined by the impulse response of two concatenated, first-order ODEs, giving g(t) = [gmax /(T - T2)] [exp(-t/T1) -exp(-t/T2)]. Voltage-dependent conductances can follow the standard HH model format, or any sort of dynamics described by a user equation. Conductances can also be made calcium-dependent.

See the URL: http://bbf-www.uia.ac.be/SOFT/NODUS_index.shtml to download Nodus.

Was this article helpful?

0 0
Apple Technologies Explained

Apple Technologies Explained

What is So Great About Apples Design Concepts? There is always a reason why consumers are drawn to a particular brand of product. Apple has the record for drawing people in, and not just Americans, why? Ease of use, thats why.

Get My Free Ebook

Post a comment