www.neuron.yale.edu

The 2006 NEURON Simulator Meeting will be held at the University of Texas, Austin,
from Friday through Sunday, May 5-7, 2006. See
http://www.utexas.edu/neuroscience/NEUR ... m2006.html
for further information.

The registration deadline (April 21) for the 2006 NEURON Simulator
Meeting is rapidly approaching! For more information see
http://www.utexas.edu/neuroscience/NEUR ... m2006.html

One of the featured speakers is Michael Hines, who will present a
tutorial and run a workshop on using NEURON to implement network
models that are distributed over multiple processors.

Tutorial Title: Parallel network simulations with NEURON

Abstract:
Parallel network management services (i.e. the ability to create
and execute network models that are distributed over multiple
processors) are now available when NEURON is configured with the
--with-mpi option. We have run extensive tests using published
network models of conductance based neurons, on parallel hardware
with dozens to thousands of CPUs. These tests demonstrate speedup
that is linear with the number of CPUs, or even superlinear (due
to larger effective high speed memory cache), until there are so
many CPUs that each one is solving fewer than ~100 equations.


Workshop Title: Implementing parallel network simulations with NEURON

Abstract:
This workshop is devoted to teaching how to transform serial network
NEURON models into a parallel program. Transformation turns out to
be fairly straightforward if the network model was originally developed
from a synapse-centric or target cell viewpoint. In other words, since
a NetCon that connects to a target cell exists only on the CPU where
the target cell exists, it is easier if one organizes the code around
the question "who projects to me?" than from the source cell perspective
"to whom do I project?". The discussion will cover important practical
and theoretical considerations, including the following:
--mpi installation and building NEURON on Beowulf clusters and other
multiprocessor systems
--how to handle random connections and random spike inputs in a way
that preserves double precision quantitative identity regardless of
number of CPUs and how cells are distributed among the CPUs
--how to measure performance

The agenda for the 2006 NEURON Simulator Meeting continues to
grow. Here are three more reasons to join us at UT Austin,
May 5-7--

Discussion: "NEURON: quo vadis?"
Moderator: Michael Hines
The agenda is to discuss future directions for NEURON, in
order to help it continue to evolve to meet the changing
needs of its users. Topics that will be considered include
strategies for improving the development process, possible
enhancements of computational efficiency and power, and
tactics for increasing functionality and ease of use.

Tutorial: "Converting morphometric data to models with the
Import3D tool"
Speaker: Ted Carnevale
One recent addition to NEURON is the Import3D tool, which was
created to simplify the task of converting detailed morphometric
data to NEURON models. Import3D can read Eutectic, SWC, and
Neurolucida classic and Version 3 files, and can export the
data directly into the CellBuilder or generate a "top level"
instance of a model. It automatically identifies and repairs
many common problems, and helps users identify other errors
that require the exercise of judgment and manual editing of
a copy of the original morphometric data.

Workshop: "Hacking NEURON."
Moderators: Ted Carnevale and Bill Lytton
The aim of this event is to exchange "power programming" tips
for getting the most out of NEURON. Discussion and specific
practical examples will range over topics such as:
● essential idioms in hoc
● organizing programs for efficiency and clarity
● combining hoc code and the GUI to exploit the strengths of both
● discovering and exploiting the single largest collection of
useful hoc code
● hacking session files
● hacking NEURON's run-time system
● custom initializations
● automating execution of simulations and analysis of results
The moderators will present many of their own tips and tricks,
and audience participation is strongly encouraged.

Register now--the registration deadline is Friday, April 21.
For more information see
http://www.utexas.edu/neuroscience/NEUR ... m2006.html

Register now for the NEURON Simulator Meeting, which
will be held May 5-7 at UT Austin--see
http://www.utexas.edu/neuroscience/NEUR ... m2006.html
The registration deadline is Friday, April 21.

Here are two more reasons to attend:

Talk: "Rule-based artificial cell (RBAC) models in NEURON"
Speaker: Bill Lytton
There are a variety of artificial cell models available in NEURON.
These have the advantage of using forward prediction of spike times
in order to avoid the overhead of integration. Additionally, they
can be used in hybrid networks with compartmental models in order
to run large models at vastly increased speed. We have developed
new artificial neuron models that use rule sets to determine firing
based on different types of inputs and several internal state
variables. In addition to speed, this approach cleanly isolates
the various synaptic and internal state variables so that they can
be manipulated in a way that is difficult in the complex
parameterizations of full multi-compartment/multi-ion-channel/
multi-synapse models. The basic rule remains the same as that of
the integrate-and-fire model: fire when the state variable exceeds a
fixed threshold. Additional rules were added to provide adaptation,
bursting, depolarization blockade, Mg-sensitive NMDA conductance,
anode-break depolarization, and others. The implementation is event
driven, providing additional speed-up by avoiding numerical
integration.

Tutorial: "Recent advances in the CellBuilder: managing models
with spatially varying parameters"
Speaker: Ted Carnevale
This tutorial will show how to use the CellBuilder, one of
NEURON's GUI tools, to construct and manage neuron models in
which one or more parameters vary with location as functions of
an independent variable. This recent enhancement graphically
supports the idiom
forsec subset for (x, 0) { rangevar_suffix(x) = f(p(x)) }
where rangevar_suffix(x) is the parameter of interest, p(x) is
a domain function over the subset, and f is any expression.
Built-in domain functions are arc (path) distance from the
soma, radial distance from a point, and distance along an axis
in the xy plane.

Three more reasons to sign up for the NEURON Simulator Meeting,
which will take place May 5-7 at the Center for Learning and
Memory, University of Texas, Austin
http://www.utexas.edu/neuroscience/NEUR ... m2006.html

1. Space is still available, but you'll have to act quickly to
take advantage of the $120 registration fee. Applications and
and payments must be received no later than Wednesday, April 26,
or they will be subject to a catering surcharge that raises the
total cost to $150.

2. All registrants will be entitled to receive a certificate
good for a 15% discount on purchase of The NEURON Book from
Cambridge University Press (regularly selling for $85). Two
copies of the book will be given away to the winners of a
random drawing of registered attendees.

3. The list of talks, tutorials, and workshops continues to grow.
Here are some of the latest additions:

Tutorial: "Using the Channel Builder"
Speaker: Ted Carnevale
The Channel Builder is a graphical interface for creating voltage-
and ligand-gated channels whose state transitions are described by
kinetic schemes and/or HH-style differential equations. Channels
constructed with this tool execute slightly faster than equivalent
mechanisms created with NMODL, the programming language used to add
new mechanisms to NEURON. Early implementations of the Channel
Builder dealt only with models in which the gating states are
continuous functions of time, i.e. a continuous system approximation
to a large population of channels with discrete states. The current
version of this tool includes many improvements, the most notable of
which may be the efficient simulation of stochastic single channel
activity. In this mode, the gating states and simulated conductance
make abrupt transitions between discrete levels, as would be produced
by the opening and closing of individual channels in a population of
countably many channels.

Workshop: "Experiments and modeling: which details are important?"
Moderator: Fernanda Saraga
In creating realistic computational models of neurons, decisions must
be made about the many experimental details that may or may not need to
be matched in the models. In addition, computational modelers often
take values for model parameters from experiments, but these values are
not absolute numbers. They depend on the specific context of how and
in what conditions those measurements were taken. Topics of discussion
will include but are not limited to: different recording paradigms, how
recording solutions effect experimental measurements, the liquid
junction potential, access resistance, and how to decide when these
details should be incorporated into the models.

Talk: "Modeling and simulation of chaotic bursting in pacemaker neurons"
Speaker: Steffen Wittmeier
Simulations of pacemaker activity using a recently developed VLSI
analog chip showed chaotic interspike intervals (ISIs) in response
to increased excitability. To confirm these results, digital
simulations were performed using a previously published conditional
pacemaker model. Near the bifurcation point between bursting
and beating the model exhibits high sensitivity to accuracy of
numerical variable time-step integration. Simulations with an error
tolerance of 10-6 for all state variables resulted in chaotic dynamics
while decreasing the error tolerance to 10-9 resulted in irregular ISIs
and obliteration of chaos. Since analog simulations always encounter
noise we added membrane noise to the system. Small amounts of noise
were sufficient to evoke chaotic behavior. We also found that the
magnitude of noise needed to transform the system from bursting into
chaotic bursting is contingent on the excitability of the cell -
consistent with the analog simulations. All digital simulations were
verified using different ODE solvers in Matlab and NEURON. Based on
these results we suggest that digital simulations of complex systems
capable of chaotic dynamics are highly sensitive to numerical
integration accuracy at the bifurcation point to chaos. Therefore
the selection of the integration method as well as the step-size or
error tolerance requires diligence.