/home/precedings/apps/precedings/shared/storage/document/2007/262/npre2007262-1

[ ]

1st INCF Workshop on Large-scale Modeling of the Nervous System

December 1213, 2006

International Neuroinformatics Coordinating Facility Secretariat

Stockholm, Sweden

Authors

Mikael Djurfeldt and Anders Lansner

Scientific Organizer

Anders Lansner, KTH, Stockholm, Sweden

Workshop Participants

David Beeman, University of Colorado at Boulder, Boulder, CO, USA
Andrew Davison, UNIC, Gif-sur-Yvette, France
Markus Diesmann, RIKEN Brain Science Institute, Wako, Japan
Rodney Douglas, UNIETH, Zürich, Switzerland
Jens Eberhard, Ruprecht-Karls-University of Heidelberg, Heidelberg, Germany
Frederick Harris, University of Nevada, Reno, NV, USA
Michael Hines, Yale University, New Haven, CT, USA
Thomas Natschläger, Software Competence Center Hagenberg, Hagenberg, Austria
Charles Peck, IBM Research Division, Yorktown Heights, NY, USA
Shiro Usui, RIKEN Brain Science Institute, Wako, Japan
Gabriel Wittum, Ruprecht-Karls-University of Heidelberg, Heidelberg, Germany
Rapporteur: Mikael Djurfeldt, KTH, Stockholm, Sweden

Supported by the INCF Central Fund and the Swedish Foundation for Strategic Research

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ ]

Contents

Executive Summary

Introduction

Concepts

3.1 What do we mean by "model"?

3.2

Model complexity

3.3

Abstraction and level of description

3.4

Detailed versus abstract models

3.5

Realism

Directions in modeling of the nervous system - scientific needs

4.1

The role of the model in current neuroscience

4.2

Top-down and bottom-up approaches

4.3

Explicitness

4.4

Large-scale models

4.5 Simulation versus emulation

4.6 Upscaling

Software for large-scale simulations

5.1 Important properties of simulator software

5.2 Diversity of simulators

5.3 Software interoperability

5.4 Accuracy of simulation

5.5 Preprocessing and specification of large-scale network models

5.6 Declarative versus procedural model description

5.7 Postprocessing and visualization

Infrastructure

6.1 The role of the INCF with regard to large-scale modeling

6.2 Verification of simulator function

6.3 Model verification

6.4 Model reproducibility

6.5 Method development in computational neuroscience

6.6 A cyber infrastructure for computational neuroscience?

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ ]

1 Executive Summary

The goal of this workshop was to survey current demands, on-

going activities, and plans relating to development of tools for

scalable neural network simulation. Areas covered included

software components for preprocessing/model setup, as well

as for storage, analysis, and visualization of results. Partici-

pants discussed the need for coordinated action in the field

with regard to model, method and tool development.

In the following summary, we present three major findings and

seven recommendations developed by workshop participants.

Findings

The workshop participants reported:

· on difficulties in reproducing simulations from pub-

lished articles

· that the current diversity of simulators creates vigor

in the field and has benefits for the validation of

models

· on the importance of facilitating software interoper-

ability and re-use of simulation software components

Recommendations

The following recommendations were made by the workshop

participants:

· Arrange an annual workshop on large-scale model-

ing with the aim of developing standards for improv-

ing reusability of software. A workshop in the same

format as the Telluride Workshops on neuromorphic en-

gineering would give an opportunity to teach large-scale

simulation technology while simultaneously presenting

an optimal environment for practitioners and students

to discuss, test, and evaluate approaches to software

interoperability.

· Implement a standard simulator test suite. 1. INCF

should promote development of a set of simulator

"benchmarks" for the purposes of verifying correctness

of computation and serving as a standard for simulator

performance. These should preferably be selected with

respect to published models not contrived by the simula-

tor developers. . INCF should coordinate and contrib-

ute to raising funding for the implementation of the test

suite by members of the labs that have developed the

simulators. . INCF should maintain web pages with the

results of the test suite for each simulator. These pages

should be updateable so that old information can be

superseded by current best practices in each simulator.

· Implement an experimental framework for connect-

ing software components. A feasibility study should be

performed regarding the possibility of on-line communi-

cation between different software modules, for example

two parallel simulators. INCF should allocate resources

for implementing a software library with the communi-

cation interface.

· Develop publication guidelines for ensuring the

reproducibility of simulations. INCF should develop

guidelines for publications with computer simulations

and publish these on the INCF homepage as a reference

for authors. INCF should also encourage publishers

to implement policies for making models available in

conjunction with publication.

· Encourage, support and fund work on method de-

velopment within computational neuroscience. INCF

should inform funding agencies about the need for

research on methods in computational neuroscience, and

on methods for large-scale simulation in particular.

· Encourage, support and fund the production of pub-

lications on concepts and techniques for large-scale

simulations. INCF should inform publishers about the

need for published work on methods within the field of

large-scale simulations and computational neuroscience

in general, and about the current growth in this area.

· Continue work on defining areas in need of support

and coordination. Workshop participants agreed that

the 1st INCF Workshop on Large-scale Modeling of the

Nervous System was successful. However, there is still

a great need for further work on coordination within the

field. INCF should therefore arrange further workshops

with the aim of defining areas in need of support and

coordination.

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ ]

2 Introduction

Computer simulation is an increasingly important tool in neu-

roscience research. As computational modeling techniques

become integrated with experimental neuroscience, more

knowledge can be extracted from existing experimental data.

Quantitative models help to explain experimental observations

and seemingly unrelated phenomena. They assist in generating

experimentally testable hypotheses and in selecting informa-

tive experiments.

Supercomputers are becoming ever-more powerful. It has even

become possible to simulate models of a scale corresponding to

substantial parts of a small mammal's neocortex with biologi-

cally detailed compartmental neuron models. At the same time,

recent progress in experimental techniques holds the promise

of supplying modelers with data of an unprecedented level of

detail. Large-scale models are becoming an essential tool in

bridging multiple levels of organization in the description and

understanding of the nervous system. However, large-scale

modeling brings many new challenges, and there is a sense in

the community that we could benefit from coordination efforts

in areas such as standardization, interoperability and verifica-

tion.

The purpose of this report is to summarize discussions and rec-

ommendations of the workshop and to answer the question of

how INCF can support the large-scale modeling community.

To this end, section surveys the current status of large-scale

modeling in neuroscience and the scientific needs. Section 5

reviews the status of current tools for large-scale modeling and

discusses topics in need of development with regard to these

tools. Finally, section discusses how INCF can support the

community directly or contribute to changes with regard to

other organizations and scientific infrastructure.

3 Concepts

In the following, some basic concepts with regard to modeling

and simulation of the nervous system will be restated and clari-

fied. This will provide a basis for a discussion of the various

current developments in the field.

3.1 What do we mean by "model"?

Mathematical models are the language of science. According

to Wikipedia, a mathematical model is an abstract model ex-

pressed in mathematical language. Further:

An abstract model is a theoretical construct that represents

something, with a set of variables and a set of logical and

quantitative relationships between them. Models in this sense

are constructed to enable reasoning within an idealized logical

framework and are an important component of scientific theo-

ries. Idealized here means that the model may make explicit

assumptions that are known to be false (or incomplete) in some

detail. Such assumptions may be justified on the grounds that

they simplify the model while, at the same time, allowing the

production of acceptably accurate solutions.

It should be remembered that a mathematical model of reality

should always be regarded as idealized in the sense above. At

least this holds true for all types of model considered in this

report.

3.2 Model complexity

The golden standard with regard to model building is well cap-

tured by words often attributed to a certain famous physicist:

"Everything should be made as simple as possible, but not

simpler." Model complexity can be measured in many ways.

For this discussion, two measures are particularly important:

1. the number of model parameters, and . the number of state

variables in the model, which will here be denoted as model

dimension.

A model is always tailored to answer a specific set of ques-

tions. For a physicist, the ideal is to pick the model with the

smallest number of parameters which will still be sufficient to

answer the scientific questions posed. This has several good

consequences:

1. A simple model can be analyzed and understood. More

complex aspects of nature can be understood through

the strategy of divide-and-conquer.

2. A simpler model expresses a simpler scientific hypoth-

esis in the sense of Occam's razor. It cuts away elements

that are irrelevant to the problem studied.

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ ]

. A model with fewer parameters is better constrained,

making it easier to falsify. A model with many param-

eters can easily be adapted to the results of any experi-

ment.

Of the two dimensions of simplicity mentioned, the first one

is most important. A model which has few parameters is more

likely to express a well defined piece of scientific knowledge

in the form of a strong, i.e. falsifiable, hypothesis, while the

lack of this kind of simplicity threatens all three of the benefits

above. A large model dimension can make the model harder to

analyze and understand but there are many techniques avail-

able for handling this kind of complexity.

In addition, it is important to make a distinction between free

parameters--parameters which are tuned by the modeler or

software to achieve a certain model behavior--and parame-

ters constrained by experimental data. A model should have

few free parameters in order not to lose benefit 3 above. Also,

all model behavior which is used to tune parameters is trans-

formed from an "output" of the model to an "input", i.e., it can

no longer be claimed as a result, or prediction, of the model.

3.3 Abstraction and level of description

The tool used to achieve a simple model is abstraction. The

system is described at a certain level and elements which are

not believed to be important for answering the scientific ques-

tions asked are taken away.

Neuroscience spans many levels of description from molecules

to behavior. But what does level mean? Churchland and Se-

jnowski (199) discuss three categories of levels. The structure

of the nervous system has many different spatial scales with

substructures such as molecules, synapses, neurons, micro-

circuits, networks, regions and systems (list slightly modified

from Churchland). We call these levels of organization. With

some good will, behavior could be added at the top of this

hierarchy.

Marr (198) described levels along a different dimension in his

levels of analysis: 1. the computational level of abstract prob-

lem analysis, . the algorithmic level, specifying a formal pro-

cedure to perform the task so that a given input will yield the

correct output, . the level of physical implementation. Marr

argued that a higher-level question was largely independent of

the levels below and could be analyzed independently of the

lower level. However, it should be noted that Marr used, to a

large extent, neurobiological considerations to constrain and

inspire his computational theories and algorithms.

Churchlands third category, levels of processing, will not be

discussed here.

For a model of the primate primary visual cortex, V1, Marr's

computational level would correspond to what computations

are being performed in V1 and why. The algorithmic level

would correspond to how the information being processed on

the computational level is represented and how the computa-

tions are carried out, while the level of physical implementa-

tion describes the actual computational elements performing

these computations.

A typical large-scale network model, with single- or multicom-

partment units, thus belongs to the level of physical implemen-

tation, since it deals with neurons and synapses. But it can still

be inspired, and constrained, from the other two levels, and can

embody principles from the "higher" levels.

So far, we have discussed levels of organization and levels of

analysis. An abstract model leaves out aspects of the descrip-

tion of reality in order to achieve simplicity. Sometimes this

means leaving out elements from a lower level of organization,

but it can also mean describing something at a higher level of

analysis. A model of visual processing in terms of filter banks

and kernels is considered more abstract than a model of neu-

ronal populations in V1. Thus, we also talk about level of ab-

straction.

3.4 Detailed versus abstract models

A detailed model is often considered the opposite of an ab-

stract model. In this report, a detailed model is defined to mean

a model which spans several levels of organization. A model

which spans the levels from networks to behavior can thus si-

multaneously be more abstract and more detailed compared to

a model restricted to a single but lower level of organization.

A model of brain imaging data which incorporates networks

of simple units can be more detailed than a statistical model of

an ion channel. Yet, because it leaves out so much of the detail

beneath the level of networks, it can also be more abstract than

the latter model.

3.5 Realism

In order to say something about reality, and in order for a cor-

responding hypothesis to be falsifiable, a model needs to be

well rooted in empirical data, i.e. formulated in a way that is

consistent with a large set of experimental data.

It should now be made clear that, just as a model can be for-

mulated on all of the levels of organization discussed in sec-

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ 8 ]

tion ., empirical data can be obtained on different levels of

organization. It should also be pointed out that models can be

formulated on higher levels of analysis (in the sense of sec-

tion .): A cognitive psychologist can retrieve behavioral data

and construct models at the algorithmic level, without a direct

reference to how these processes are physically implemented

in the brain. Brain imaging retrieves data at the levels of net-

works, regions and systems. Electrophysiology collects data at

many levels of organization and this data can be used to con-

struct models at the level of physical implementation.

A model based on data from a lower level of organization, or

formulated in terms of elements from multiple lower levels of

organization, is not necessarily more realistic than data formu-

lated and rooted at a higher level. Rather, a model is realistic

if it is well rooted in empirical data at the given level, if its pa-

rameters are well constrained by these data, and if it correctly

predicts data which has not been used to tune the model.

Not only the parameters of a model, but also the equations,

represent assumptions about reality. Some models are based on

first principles, that is, their equations are established laws of

physics. Models based on first principles, or models with equa-

tions which can be derived from the laws of physics, can be

trusted to a larger extent than other models, because they rely

less than other models on assumptions in the form of peculiari-

ties of the model equations or fitting of parameters.

4 Directions in modeling of

the nervous system - scientific

needs

4.1 The role of the model in current neuroscience

This section contains a brief discussion of the various roles

that models currently have in neuroscience. As stated above,

models are the language of science. We use models to

· formulate hypotheses regarding the function of the ner-

vous system

The activity of formulating a hypothesis in terms of a model

requires collection of experimental data. It is often discovered

that crucial data are missing. In this respect a model could be

considered

· a tool to identify what we don't know

Often, hidden contradictions and inconsistencies are revealed

during the formulation process, and it happens that models

don't yield expected results so that a further role of the model

· validating self-consistency of the description of a phe-

nomenon or function

If the confidence in the model is strong but the predictions dif-

fer from experiment the model can be used to

· falsify hypotheses

If phenomena in the model are unexpected or unobserved a

model can

· suggest new experiments

Finally, a model can be used as a

· platform for integrating knowledge

unifying experimental data from many sources in a consistent

manner.

4.2 Top-down and bottom-up approaches

The top-down approach to modeling most often means using

hypotheses on the computational or algorithmic levels as a

starting point when approaching the formulation of a model at

the level of physical implementation. These functional hypoth-

eses guide the formulation of the model at the implementation-

al level in terms of what elements are included in the model

and which experimental data are considered important. A func-

tional hypothesis can also complement experimental data in

the sense of giving additional constraints. For example, if we

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ 9 ]

have limited experimental data on the functional connectivity

between inhibitory and excitatory connectivity, an additional

functional constraint would be that the activity in the network

must not be allowed to grow in an uncontrolled manner.

A top-down approach can also start with an abstract model,

neglecting detail below the level of organization at which the

model is formulated. Succeeding models can then increase the

amount of detail, enabling them to account for a larger set of

experimental data. For example, a connectionist, rate-based

model can be developed into a spiking network model, fol-

lowed by a Hodgkin-Huxley style model. However, it should

be noted that models are more typically developed outside of

such sequences, at a specific level of abstraction and detail

suitable to the scientific questions posed.

The bottom-up approach, in contrast, means using the level of

physical implementation as a starting point with the hope of

capturing function. For example, what does the anatomy of the

cerebral cortex mean? If we can, from the physical level of

synapses, dendrites, neurons and networks, identify computa-

tional primitives of the cortex such primitives can be abstracted

and we can move up one level of analysis. This is one goal of

projects like DAISY (Kennedy, 00), FACETS (Meier, 00)

and Blue Brain (Markram and Peck, 00).

As with the top-down approach, a bottom-up approach can be

concerned with levels of organization. In this case, it means to

take a lower level of organization as a starting point for under-

standing the higher level.

In practise, the approach of a modeller is usually neither pure-

ly top-down nor purely bottom-up, as was already evident in

Marr's work. Also, in the bottom-up approach, the focus of

experiments and the choice of elements to include in the model

is largely guided by functional hypotheses.

4.3 Explicitness

We define model explicitness to mean the degree to which the

model is isomorphic with reality, or, how directly state vari-

ables of the model can be mapped to empirical data.

The degree of detail in a multi-compartment, Hodgkin-Huxley,

model of a neuron aids in making this type of model transpar-

ent in the sense above. During intracellular recording of a neu-

ron, it is often possible to block a subset of the ion channels in

order to directly measure the current of another channel subset,

or, in order to study the effect on cell behavior. The explicit-

ness of the Hodgkin-Huxley formalism then makes it easy to

perform a corresponding manipulation in the model. It is also

easy to identify and display individual currents in the model.

In comparison, an integrate-and-fire model is less transparent.

While recent versions (e.g. Brette and Gerstner) of this type of

model can faithfully reproduce a spike train, some state vari-

ables correspond to the phenomenological effect of the coor-

dinated action of multiple channel types in the real neuron. In

this case, it is easy to compare the spike trains, but it is not

as easy to map the dynamics of the model to the individual

currents of the real neuron. On the other hand, the simplicity

of the integrate-and-fire model makes it easier to analyze and

understand.

Another aspect of explicitness is that it supports the role of

the model as a platform for integrating knowledge. A transpar-

ent model is more likely to connect well to a wider range of

experimental data, even data which were not targeted when

constructing the model.

Workshop participants commented that, in the past, there have

been many sacrifices in explicitness in order to achieve per-

formance. Early connectionist models used simple rate-coding

units, enabling the simulation of models of networks on the

personal computers of the 1980s, while Hodgkin-Huxley type

models could only be simulated one neuron at a time. Today,

we can simulate large networks of neurons with thousands of

compartments.

4.4 Large-scale models

For the purposes of this report, a large-scale model is a model

with a high dimension, i.e. a model with a large number of

state variables (on the order of hundreds of millions or more).

Thus, a detailed model of one cortical column can be large-

scale while an abstract model of a large network encompassing

multiple columns is not necessarily large-scale.

4.4.1 Integrate-and-fire models

The classical integrate-and-fire model (MacGregor and Oliver,

19; Tuckwell, 1988) has one state variable per neuron, rep-

resenting the membrane potential. It is basically a linear leaky

integrator with a voltage threshold and a reset mechanism. The

main advantage of this type of model compared to Hodgkin-

Huxley type models is its simplicity. For example, it has far

fewer parameters, while it still captures essential features of

the neurons. Because it has fewer parameters it is easier to

adapt this type of model to experimental data. In this sense, it

is easier to achieve a certain level of realism with an integrate-

and-fire model than with a Hodgkin-Huxley type model, while,

with more effort and more data, the latter model can reach an

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ 10 ]

even higher degree of realism. Its simplicity also makes it pos-

sible to develop automated procedures for extracting param-

eters from data (Jolivet et al., 00; Keat et al., 001; Paninski

et al., 00). Because of the low dimension and mathematical

tractability, it is easier to analyze and understand this type of

model. Finally, the simulation of this type of models require

fewer computational resources making it possible to simulate

larger networks given the same hardware. For examples of

large integrate-and-fire-based network models, see Mehring

et al. (00); Aviel et al. (00); Tetzlaff et al. (00); Kumar

et al. (00); Morrison et al. (00a).

The integrate-and-fire paradigm has recently been developed

in three directions (Brette and Gerstner, 00):

· the addition of a quadratic or exponential term, yield-

ing a smooth spike initiation zone (Latham et al., 000;

Fourcaud-Trocme et al., 00);

· the addition of a second state variable, enabling model-

ing of subthreshold resonances or adaptation (Izhikev-

ich, 00; Richardson et al., 00);

· using active conductances to model synaptic inputs

(Destexhe et al., 00).

4.4.2 Hodgkin-Huxley models

A network model with multi-compartmental Hodgkin-Huxley

type units is more detailed than its integrate-and-fire counter-

part. It is more complex, both by having more parameters and

a larger model dimension. As has been discussed in section

..1, this requires more labor to determine parameters from

experimental data. Sometimes, not all data is available so that

parameters need to be determined more indirectly. The disad-

vantage is that this turns an output of the model into an input

(.). For example, if we tune the conductance of a K

chan-

nel in order to obtain the correct time course of an AHP, instead

of measuring this conductance, we can no longer claim that our

model predicts a correct AHP. In this case, however, the kind

of predictions which are of interest in a network model appear

at another level of organization within the model.

In section ., the explicitness of HH-type models was dis-

cussed. The higher level of detail allows this type of model

to be connected to a wider range of experimental data. The

presence of ionic currents allows for comparatively easy mod-

eling of pharmacological manipulations. The D extent of a

compartmental model allows for the synthesis of EEG and LFP

signals (Einevoll et al., 00).

4.4.3 The bottom-up approach to the cortical
column

Regardless of whether we use integrate-and-fire or Hodgkin-

Huxley type units in a network model, an important set of

parameters that currently largely lacks an experimental basis

is the set of connectivity parameters. Data from, for example,

Thomson et al. (00) and Binzegger et al. (00) gives the

statistics of connectivity between pairs of cell types. This type

of data has led to the use of random Gaussian (e.g. Brunel)

or random uniform (e.g. Haeusler and Maass) connectivity in

network models, consistent with such statistics. However, it is

reasonable to assume that the microcircuitry of a column has

more structure than that. Also, there is very limited data on the

structure of long-range connectivity.

The Blue Brain project (Markram and Peck, 00) aims to col-

lect data on individual cells, for example acquisition of cell

morphology through cell labeling and -photon microscopy,

and then using database techniques and specialized software to

reconstruct a virtual column. The superposition of reconstruct-

ed cells in D-space may give additional constraints needed to

get a more complete picture of microcircuitry. The aim is also

to simulate a large-scale network model of a complete column

with multicompartmental Hodgkin-Huxley-type units without

reference to functional hypotheses about the network. Thus,

the approach is essentially hard-core bottom-up.

Workshop participants commented on the difficulty in experi-

mentally determining the existence of a synaptic contact be-

tween cortical neurons since axonal processes can pass close to

the dendritic processes of neurons without forming a synapse.

In essence, the only way to determine anatomically if there is

a contact is by looking at it with an electron microscope. It is

also possible to determine the existence of a connection elec-

trophysiologically by recording from pairs of neurons (e.g.,

Thomson et al.).

One particularly interesting development with regard to the ac-

quisition of connectivity data was discussed at the workshop.

Denk and Horstmann (00) have developed a method called

"serial block-face scanning electron microscopy" or SBFSEM.

A microtome is placed in the chamber of a scanning electron

microscope. The face of the tissue sample is scanned and 0

0 nm slices cut away, generating stacks of thousands of im-

ages from which a bulk D volume can be reconstructed. The

acquired data has enough resolution to trace thin axons and

identify synapses. This method holds the promise of geometri-

cally reconstructing an entire neocortical minicolumn and ex-

tracting its circuitry.

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ 11 ]

Workshop participants concluded that the acquisition of data

at multiple levels of organization leads to a new scale of proj-

ects--a development which parallels the study of the genome

and studies in particle physics. There will be new kinds of

challenges associated with industrial-scale data acquisition in

projects of thousands of man-years.

4.4.4 Combining the top-down and bottom-up
approaches

A survey of existing literature in computational neuroscience

shows that the pure bottom-up approach to understanding net-

work function is very rare. Part of the explanation is that miss-

ing empirical data need to be complemented with functional

hypotheses in order to make progress possible. But a top-down

approach may have importance in other ways than replacing

lacking knowledge. During the workshop, the question arose

whether a correctly implemented, detailed computer replica of

the cortical column would by itself say very much about net-

work function. Because a detailed model can be complex and

hard to analyze and understand, modellers tend to see func-

tional hypotheses and abstract models as a necessary comple-

ment in dissecting cortical function. It should still be noted that

a computer replica is much more accessible to experimentation

than the living tissue in the sense that any set of state variables

can be logged and an arbitrary set of variables can be simulta-

neously perturbed in a precise fashion.

An example of how a top-down approach can be combined

with the bottom-up approach is given by the model of Lun-

dqvist et al. (00) (see Djurfeldt et al. for a large-scale ver-

sion of the model). The model is mainly designed to target the

question: Is neocortical microarchitecture consistent with the

hypothesis of attractor memory network function? Here, most

parameters of the neuron models are determined from experi-

mental data in a bottom-up manner. However, the connectivity

parameters are determined by combining a long-range connec-

tivity structure required for attractor memory network function

with the currently existing empirical constraints on connectiv-

ity mentioned in section ...

Another aspect of this model, and most or all other network

models, is that the parameters of a neuron type are replicated

over the population of model neurons, with or without random

perturbation, in a crystal-like manner. This means that even if

a large-scale model of this type has a large model dimension, it

can still have a comparatively small number of parameters and,

thus, be simple in the important sense (c.f. section .).

4.4.5 Volume simulation

Until now, Hodgkin-Huxley type models have represented the

most basic level of organization at which we simulate neurons

and circuits, with the exception of hybrid models also includ-

ing biochemical processes inside the cell. Data from the SBF-

SEM method mentioned in section .. opens up the prospect

of a full D volume simulation of a cortical column. During

the workshop, Gabriel Wittum presented initial attempts in this

direction at the level of a single cell. In the Hodgkin-Huxley

approach, the neuron is modelled as an electrical circuit. Here,

instead, the D volume in which the neuron is embedded is de-

scribed in its entirety by partial differential equations (PDEs)

and simulated using the solver µG (Bastian et al., 199; Wit-

tum, 00).

A model of the D volume can be based on

first principles in

the sense discussed in section .. In this case, the model is

based on Maxwell's equations which describe the dynamics of

the electromagnetic field.

4.4.6 Simulation of growth processes

In order to fully understand the cortical architecture, it is neces-

sary to understand the development and growth processes from

which it results. Within the DAISY project, initial steps are

currently taken to simulate the migration of neuroblasts. This

adds a requirement on the simulator which is not yet fulfilled

by standard neuron simulators such as Neuron and Genesis:

there is a need to quantize space. Simulation tools are being de-

veloped within DAISY to meet this demand. The solver µG is

based on a communication layer, DDD (Dynamic Distributed

Data), which allows computational loads to migrate within a

parallel computer during simulation. This layer could also be a

suitable substrate for the simulation of growth processes.

4.5 Simulation versus emulation

Alan Turing used the term simulation in a very specialized

sense. The term referred to the simulation by a digital com-

puter of a subject discrete-state machine, defined by a set of

state transitions, inputs and outputs.

In computational neuroscience, the term usually refers to the

computation of the numerical approximation of a solution over

time to the equations of a mathematical model. When per-

formed on a digital computer, such computations are subject

to the limitations of the computer. For example, some quantum

mechanical processes can not be simulated on a digital com-

puter. However, such limitations do not seem to be of any rele-

vance to current computational models of the nervous system.

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ 1 ]

By an emulation, we refer to a model of the nervous system

in the shape of a physical realization, for example in terms of

an electronic circuit. The projects DAISY and FACETS both

implement VLSI chips emulating neural circuits.

During the workshop, participants commented that while most

simulations are slower than real-time (time simulated is shorter

than the wall-clock time required to compute the solution), it

is possible to construct an artefact which can emulate a neural

circuit in real-time. The important consequence is that the arte-

fact can interact with the environment in real time and react to

a continuous flow of events occurring asynchronously. In some

cases this is also achievable with a simulation on a digital com-

puter, as exemplified by the dynamic-clamp technique (Sharp

et al., 199) or the goal of the Blue Brain project to simulate a

cortical column on a super-computer in real time.

4.6 Upscaling

A model of the bulk D volume of neural tissue using PDEs

(section ..), if well-rooted in empirical data, can be consid-

ered more realistic than the Hodgkin-Huxley model. It is based

on first principles (sections 4.4.5, 3.5) while the Hodgkin-

Huxley model is partly based on simplifying assumptions and

curve fitting. This means that we can use the 3D volume model

to validate the multicompartmental Hodgkin-Huxley model.

Gabriel Wittum reported on simulations of a single cell where

ephaptic interactions could be observed between two of the

dendritic processes of the cell itself. Such a phenomenon

would not arise in the Hodgkin-Huxley model. It is also clear

that the surrounding interstitial fluid, neuropil, and dendritic

processes have substantial effects on the electric behavior of

the cell membrane, diverging from the Hodgkin-Huxley mod-

el. Clearly, however, the D volume simulation is not a good

replacement for the Hodgkin-Huxley model, because it is com-

putationally heavier. The question then arises what alternatives

we have to the HH model.

An important answer is given by the D model itself: There

exist mathematical techniques for deriving a model at a coarser

scale from a model at a finer scale. This methodology is called

upscaling (Eberhard et al., 00). Through upscaling tech-

niques it might be possible to derive a candidate model which

might serve as a better replacement for the Hodgkin-Huxley

model.

5 Software for large-scale

simulations

5.1 Important properties of simulator software

When judging software for large-scale simulation, there are

many criteria that need to be examined, some of which will be

mentioned in this section. Brette et al. (00) reviews existing

tools for simulation of networks of spiking neurons.

· Model types supported What neuronal/synaptic/plas-

ticity models can be simulated?

· Accuracy Does the simulator give correct results?

Recent work (Brette, 00; Rudolph and Destexhe,

00; Morrison et al., 00b) has presented methods for

determining spike times in a simulator more precisely,

and has shown that this can have effects on discharge

statistics and temporal precision in resolving synaptic

inputs. This is discussed further in section . below.

· Scalability In general, it is difficult to write efficient

parallel implementations. How does speedup (simula-

tion time divided by simulation time on one processor)

scale with the number of processors used? Ideally it

should grow linearly. How does simulation time scale as

a function of the size of the model?

· Documentation How good is the documentation?

· Support How quickly will the developers respond to

bug-reports or feature requests?

· License and availability of source code In a research

environment, it is an advantage to have the source code

for the simulator available, and to have permission to

modify it. This is guaranteed for all software covered by

the GPL license from the Free Software Foundation and

some related licenses.

· Adaptability How easy is it to adapt the simulator to

your purpose? How easy is to to add new mechanisms?

· Portability Does it run on my preferred platform?

· Interoperability How easy is it to collaborate with oth-

ers using a different simulator? This is discussed further

in section . below.

· Is there a graphical interface?

· What analysis/post-processing tools are available?

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ 1 ]

5.2 Diversity of simulators

During the workshop, it was discussed whether it is sensible

for the community to split efforts into the large and growing

set of neuron simulators available today rather than focussing

on one or a few tools. However, it was also noted that there is

currently a strong ongoing development of simulation technol-

ogy, and that simulators tend to have unique strong points not

shared by others.

The workshop also identified the reproducibility of models as

a major problem (see section .). The diversity of simulators

might allow for simulating the model on a different platform

from that on which it was originally developed, thereby verify-

ing the reproducibility of results. This leads to the first major

finding:

Workshop participants agreed that the current diversity of

simulators creates vigor in the field and has benefits for the

validation of models.

5.3 Software interoperability

Given the diversity of simulators, it is important to find ways

to share the gains produced by the efforts, of both developers

and users, put into individual simulators. During the workshop

some approaches to simulator-independent modeling environ-

ments were discussed, which allow a model to be simulated

using more than one simulation engine. This is important for

verification of simulator accuracy, for the reproduction, test-

ing and extension of published models, and for collaboration

between modellers using different simulation tools. The ap-

proaches discussed were graphical environments, declarative

model specifications using XML, and procedural model speci-

fications using an API implemented in the Python program-

ming language (see sections .. and .). Another conclu-

sion was that more effort should be put into the development of

such simulator-independent environments (`meta-simulators').

Thus, the second major finding of the workshop was:

Workshop participants agree upon the importance of fa-

cilitating software interoperability and re-use of simulation

software components.

5.3.1 Modularity

The practise of dividing software into modules with well-de-

fined roles has many advantages. It eases development and

Michael Hines noted that "The reason why we keep rein-

venting the wheel is that we haven't got it quite round yet."

increases maintainability of the code. If such modules have

well-defined interfaces, modules can be re-used in other cir-

cumstances. Such an interface can be in the form of

1. an application programming interface (API), enabling

a module in the form of a compilation unit or a library

to be linked into an application. This includes the defini-

tion of data structures required to pass information

through the interface.

. a communication interface, enabling modules in the

form of processes to communicate while running simul-

taneously on the same or on different machines

. a file format, allowing the output of a module in the

form of an application to be read as input to another

application. Here, "input" can be a model specification.

In this case, the interface takes the shape of a model

specification language.

As an example of the first form of modularity, a simulator can

be divided into a simulator kernel, responsible for the dis-

tribution and allocation of data structures over a cluster, for

building the model on the nodes, and for performing the com-

putations during a simulation, and other modules required for

module specification, input and output. The simulator kernel

can be further divided into a solver, with the sole responsibility

of performing computations, and modules required for alloca-

tion, distribution etc.

Workshop participants discussed the possibility of on-line in-

teraction between simulators (see section ..). This would

entail modularity of the second type.

An example of the third type of modularity is neuroConstruct

which is a software application for creating D models of

networks of biologically realistic neurons through a graphi-

cal user interface (GUI) (.1.). neuroConstruct can import

morphology files in Genesis, Neuron, Neurolucida, SWC and

MorphML formats for inclusion in network models and can

generate model specification files for Genesis and Neuron. Ef-

forts put into developing neuroConstruct further will thus ben-

efit both the Genesis and Neuron communities. Note, though,

that the choice of two output formats is forced by the current

lack of a standard format for model description. This will be

discussed further in section ...

David Beeman presented how the next generation of Gene-

sis, Genesis (..), aims to foster collaborative modeling

through a rich set of interfaces.

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ 1 ]

We cannot gain fully from a modularization of simulation soft-

ware until we have developed standard interfaces between

software components.

When facing the need to develop standards, there is the pos-

sibility of assigning the task to a standardization committee.

Workshop participants agreed, however, that there are other

alternatives which might be preferable. It was suggested that a

workshop should be arranged with development of interoper-

ability as one goal.

Participants reported on good experiences with the form of the

Telluride workshops (van Schaik et al., 00) which includes

background lectures from leading researchers, practical tuto-

rials from state-of-the-art practitioners, hands-on projects in-

volving established researchers and newcomers/students, and

special interest discussion groups proposed by the workshop

participants. The INCF workshop gives the following recom-

mendation:

Arrange an annual workshop on large-scale modeling with

the aim of developing standards for improving reusability

of software. A workshop in the format of the Telluride Work-

shops on neuromorphic engineering would give an opportunity

to teach large-scale simulation technology while simultane-

ously presenting an optimal environment for practitioners and

student to discuss, test, and evaluate approaches to software

interoperability.

5.3.2 On-line interaction between simulators

As was discussed in section ., different simulators have

different strengths. Workshop participants observed that cur-

rent development is moving in the direction of simulation of

large systems of networks. The situation could arise that one

simulation framework is most suitable for one part of the sys-

tem while another framework is needed for another part. For

example, a retina model could provide input for a model of

LGN/PGN/V1, or, a medium-sized network of detailed struc-

turally realistic neurons could interact with a very large net-

work of simple integrate-and-fire point neurons. In the former

example, one solution would be to perform the simulation in

batch mode, letting the retina model generate spikes and save

the spike times to file. Such spike files can then be read by the

second simulator. This would be an example of the third type

of modularity in section ..1. In the latter example, this is

not possible due to the need for bi-directional interaction. This

situation is resolved if the simulators can transfer spikes on-

line through a communication interface.

Even in cases where batch mode simulation is possible, it may

be desirable to let such a simulation interact with the environ-

ment in real time. Another argument for preferring on-line

communication between simulators over batch mode simula-

tion is if the amount of data generated is of such a large mag-

nitude that it is undesirable or impossible to store intermediate

data in files.

If both simulators are run on the same parallel computer, some

further coordination may be required with regard to allocation

of nodes, initialization of MPI, etc. This could be achieved if

the communication interface has the form of a communica-

tions library, providing services like initialization of MPI and

the option of external communication. There is also a need for

a naming or addressing mechanism to allow for flexible con-

nectivity between modules through multiple communication

links.

The development of such an interface may or may not result in

a software component that will be generally adopted, but has

value in itself as an exploration of the concept. The workshop

gives the following recommendation:

Implement an experimental framework for connecting soft-

ware components. A feasibility study should be performed re-

garding the possibility of on-line communication between dif-

ferent software modules, for example two parallel simulators.

INCF should allocate resources for implementing a software

library with a communication interface.

5.3.3 Common specification language

Different simulator environments have different ways of speci-

fying a model. For example, Neuron and Genesis have distinct

specification languages. The translation of a model description

from one environment to another can entail substantial effort.

This is especially true in the case that environments use dif-

ferent formalisms. If the differential equation for the activa-

tion factor in the Hodgkin-Huxley model of a channel has dif-

ferent forms in the two environments, the translation of the

model will even involve finding a new set of parameters. This

situation creates barriers between laboratories using different

simulators, makes it harder for one laboratory to freely choose

the tool suitable for the problem at hand, and threatens the re-

producibility of scientific results. A common, standard speci-

fication language, supported by all simulators, would alleviate

such problems and increase the utility of model repositories

such as ModelDB (..1) and databases such as NeuronDB

(..).

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ 1 ]

NeuroML (..1) was discussed as one viable candidate for a

model description standard. NeuroML is a collection of proj-

ects with the aim of developing standards for specification of

neuroscience models in XML. It is organized into four levels

of scale, where each succeeding level extends the features of

the language. The first level covers the neuronal level of or-

ganization while the last level (level ) covers all levels of or-

ganization from biochemical networks to systems. It includes

the XML schemas MorphML, ChannelML and NetworkML.

Workshop participants noted that NetworkML is currently the

least developed schema and needs input from simulator devel-

opers.

PyNN, discussed in section .., was also proposed as a can-

didate for a standard model description.

5.3.4 Common scripting language

A scripting language is an interpreted language which is used

to control an application. It is often embedded in the sense that

much of the functionality of the application is available as

function or procedure calls in the language--a language bind-

ing. If it is possible to add new functionality to the application

in terms of code written in the scripting language, it is also

called an extension language. The use of a full-fledged general

purpose programming language as scripting/extension lan-

guage has emerged as a powerful concept. Scripting languages

such as Emacs Lisp, TCL, Matlab and Python have each given

rise to prolific user communities and rich sets of software tools

and libraries, and can provide a backbone in a framework with

multiple modules.

Within the FACETS project (Meier, 00), Andrew Davison

has proposed the PyNN framework (..) as a standard script-

ing language binding for neuronal network simulators. This

abstracts differences between simulators and provides a com-

mon way to specify models and run simulations. The goal is

that simulation scripts in PyNN for simulator A will run on

simulator B without modification.

PyNN is based on the Python programming language and in-

cludes the development of an API and the binding to individual

simulation engines. The API has two parts, a low-level, proce-

dural API, and a high-level, object-oriented API. The low-level

API can be used for small networks, and gives more flexibility

than the high-level API. The high-level API hides details and

book-keeping, and is intended to have a one-to-one mapping

with NeuroML, i.e. a population element in NeuroML will cor-

respond to a Population object in PyNN, etc. Another require-

ment for a common scripting language is standard cell models.

PyNN translates standard cell-model names and parameter

names into simulator-specific names.

The use of Python in PyNN results in a free, Matlab-like en-

vironment with tools for data analysis, plotting, mathematical

libraries, etc., leveraging the efforts of the Python user com-

munity.

5.4 Accuracy of simulation

5.4.1 Combining off-grid spike events with time-
driven global scheduling

Very large networks of spiking neurons can be simulated ef-

ficiently in parallel under the constraint that spike times are

bound to an equidistant time grid. Within this scheme, the sub-

threshold dynamics of a wide class of integrate-and-fire type

neuron models can be integrated exactly from one grid point

to the next. However, the loss in accuracy caused by restrict-

ing spike times to the grid can have undesirable consequences,

which has led to interest in interpolating spike times between

the grid points to retrieve an adequate representation of net-

work dynamics. Markus Diesmann demonstrated during the

workshop how the exact integration scheme can be combined

naturally with off-grid spike events found by interpolation

(Morrison et al., 00b). By exploiting the existence of a mini-

mal synaptic propagation delay, the need for a central event

queue is removed, so that the precision of event-driven simu-

lation on the level of single neurons is combined with the ef-

ficiency of time-driven global scheduling. Further, for neuron

models with linear subthreshold dynamics, even local event

queuing can be avoided, resulting in much greater efficiency

on the single neuron level. A measure of the efficiency of net-

work simulations in terms of their integration error shows that,

for a wide range of input spike rates, these novel techniques are

both more accurate and faster than standard techniques.

5.5 Preprocessing and specification of

large-scale network models

Section .. discussed the bottom-up approach to large-scale

modeling of the cortical column using data from the SBFSEM

method. For a D volume simulation (section ..), a corre-

sponding D volume of dielectric properties must be generated

as part of the model specification. For a more traditional simu-

lation based on compartment model neurons, the data must be

analyzed with respect to cell morphology and a list of network

connections be generated.

Section .. discussed a combined top-down/bottom-up

approach where the connectivity is based on functional hy-

potheses regarding cortical function. The Brunel (000) and

Haeusler and Maass (00) models also belong to this type. In

such models, there is a large contrast between the seemingly

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ 1 ]

complex connectivity of the simulated model and the simple

ideas upon which the connectivity is based. Usually the con-

nectivity is expressed by a serial algorithm that sets up connec-

tions one by one. On a parallel machine, several simulators use

the approach of running the serial algorithm in each process

independently and simply ignoring attempts to set up connec-

tions not belonging to the local process. The advantage is that

the complexity of parallelism is hidden from the user. How-

ever, this approach does not scale well. Djurfeldt et al. (00)

presents an approach which scales well at the same time as

expressing the connectivity in a form that preserves the under-

lying ideas explicitly in the program code. Projections between

neuronal populations are described by iterators representing

infinite connection matrices. Only the relevant finite piece of a

matrix is used to connect the populations during model setup.

A basic set of parameterized matrices can be combined through

a connection matrix algebra to form new connectivity struc-

tures.

A third way to specify networks is to generate them based on

empirical structure parameters. Regarding automatic genera-

tion of realistic, large-scale neural networks, there is a need to

distinguish between generation and growing of cells. Grow-

ing of cells might depend on local or global parameters in the

network. Further, an open question is how to specify and gen-

erate spatial connectivity. The easiest way of generating con-

nectivity is using experimental data about the localization of

different neuron types in the brain. This approach is commonly

used in the available software tools. During the workshop, Jens

Eberhard presented an assessment of existing tools and evalu-

ated features such as interfaces and used formats, "large-scale"

capability, expandability, and postprocessing/visualization.

The following tools were evaluated: L-Neuron (.1.1), Arbor-

Vitae (.1.), A-Cell-D, NeuroConstruct (.1.), and NeuGen

(.1.). Based on the features mentioned above, only Neuro-

Construct and NeuGen could be recommended for the specifi-

cation of large-scale network models. These tools are capable

of generating detailed three-dimensional neuron geometry for

networks up to about 10000 cells.

5.6 Declarative versus procedural model

description

One of the discussions during the workshop concerned the

relative merits of declarative versus procedural model speci-

fication. Model definitions based on NeuroML are basically

declarative, although it is in principle possible to specify algo-

rithmic elements in XML. In contrast, a model description in a

Python-based scripting language is often procedural.

A declarative description is often easier to read and understand

than a procedural description. It also usually gives a software

tool more leeway for internal optimizations.

However, workshop participants commented that, in a cer-

tain respect, a declarative description is "un-human". This is

because it throws away structural knowledge about the prob-

lem already available to the researcher. This knowledge is ex-

pressed in algorithms which can be close to the mathematical

formulation of the model. We should look for a good high-

level description that allows a human reader to see from the

code what the point is.

5.7 Postprocessing and visualization

The kind of large-scale simulations which are the subject of

this workshop have, by definition, a large number of state vari-

ables. When the number of variables is in the order of hundreds

of millions or more, the problem how to visualize network ac-

tivity becomes important. One of the tools discussed during

the workshop was DAVIS (..1). However, there seems to

be a great need of further tool development in this area. For

example, for sufficiently large data volumes, the visualization

tool needs to be parallelized.

A large-scale model spans many levels of organization, and,

thus, needs to connect to experimental data on many levels.

There is, therefore, a need for tools which can transform logged

data from model state variables, such as cell membrane volt-

ages, into synthetic versions of common brain imaging meth-

ods such as fMRI, VSD and EEG.

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ 1 ]

6 Infrastructure

6.1 The role of the INCF with regard to

large-scale modeling

There was a consensus among workshop participants that the

field of large-scale modeling of the nervous system has reached

a level of maturity where it would benefit from coordination

efforts, for example with regard to simulation tools, databas-

es and publication of results. Various specific ways in which

INCF could assist such efforts were discussed, some of which

are summarized in the following sections.

One of the roles suggested for the INCF was to develop an

expert definition of where things stand in the field and iden-

tify complementarity between research groups, i.e. make an

inventory of what is available in different places, define areas

in which it is important to achieve progress, and facilitate this.

The workshop gives the following recommendation:

Continue work on defining areas in need of support and

coordination. Workshop participants agreed that the 1st INCF

Workshop on Large-scale Modeling of the Nervous System

was successful. However, there is still a great need for further

work on coordination within the field. INCF should therefore

arrange further workshops with the aim of defining areas in

need of support and coordination.

It was also suggested that the INCF could have a staff of ex-

perts available who understand various software applications

and components fully, and know how to run tools on a refer-

ence set of problems. This could provide a resource both with

regard to development of software interoperability (c.f. section

.) and teaching.

With regard to databases and platforms (see section .), it was

suggested that the INCF should promote and coordinate data

archives. What data are there? What is missing?

Workshop participants reported on a general feeling in the

modeling community that we need and can achieve synergy ef-

fects if money were to be allocated for people to guide the de-

velopment of software interoperability and common platforms.

The comment was also made that administrators of funding

agencies would surely recognize this need if it was properly

communicated to them. It was concluded that there is a com-

munication problem between researchers and funding agen-

cies in this respect and it was suggested than the INCF could

aid in this communication, for example vis-a-vis the European

Union. More generally, it was suggested that the INCF could

assist in networking people in the field.

6.2 Verification of simulator function

In science, an individual experiment carries little weight, but

when the result is reproduced in other laboratories, by different

people, in slightly different ways, this gives strong validation.

One of the major discussion topics during the workshop was a

parallel need for validation of simulator function.

It is an established software engineering fact that any suffi-

ciently complex program does contain errors, regardless of the

quality of the developer team. For a simulator, there can be

errors at the level of a model built in implicitly in a simula-

tor--e.g. the Hodgkin-Huxley neuron model, at the level of the

numerical method used, and at the level of its implementation.

Furthermore, different simulators can have different accuracy

and vary in their efficiency.

Section . discusses the diverse set of existing neuron simu-

lators. One of the ways in which to draw advantage from this

diversity is to use the fact that different simulators are pro-

grammed differently to cross-validate simulators: If simula-

tion results can be reproduced using a differently programmed

simulator, this gives a relatively strong verification of the cor-

rectness of both simulators.

For such cross-validation to work well, there is a need to agree

on a common set of simulator "benchmarks". Such a frame-

work could also make possible comparisons of accuracy and

efficiency. This could work as an inspiration and driving force

for simulator developers, and help in focussing development.

Without quantitative comparison data it is difficult for devel-

opers to know what is practically feasible.

If a benchmark suite is published on web pages maintained by

the INCF this would be of great community value. It is impor-

tant that such benchmarks be with respect to published models

not contrived by the developers. Workshop participants men-

tioned cases where essentially good methods were neglected

because they had poor performance in artificial benchmarks.

This experience parallels that in software engineering, where

the trend during the previous two decades has been from arti-

ficial contrived benchmarks towards more application-like. It

is also important that the data on the web pages be updateable

so that old information can be superseded by current best prac-

tices in each simulator.

Examples of possible benchmarks are the DeSchutter Purkinje

cell model (Schutter, 1998) and the model of Vogels and Ab-

bott (00). ModelDB (..1) contains 80+ such published

models from which a selection could be drawn. The workshop

gives the following recommendation:

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ 18 ]

Implement a standard simulator test suite. 1. An INCF

staff member should propose a set of simulator "benchmarks"

for the purposes of verifying correctness of computation and

serving as a standard for simulator performance. These should

preferably be selected with respect to published models not

contrived by the simulator developers. . INCF should allocate

funding for and coordinate the implementation of the test suite

by members of the labs that have developed the simulators. .

INCF should maintain web pages with the results of the test

suite for each simulator. These pages should be updateable so

that old information can be superseded by current best prac-

tices in each simulator.

6.3 Model verification

In section 6.2, the verification of the simulation tools was dis-

cussed. During the workshop there was also a discussion about

the verification of models and the reproducibility of modeling

results.

A crucial point for a simulation is: How do you know that what

you're simulating is correct? As was discussed in section .,

the model needs to be well rooted in empirical data. This can

be achieved if test problems are chosen with care. Such prob-

lems should deal with basic traits of the domain under study

and be chosen so that many aspects of the problem are easily

accessible through experimental measurement. The test prob-

lems can validate the simulation technique so that simulations

can then be extrapolated to other problems with some degree

of confidence. The comment was made that some areas in the

brain are easier to work with, with regard to model validation,

than others. It is especially important to look at a region with

well-defined inputs and outputs, such as, for example, the bar-

rel cortex.

Section 3.5 also discussed how models based on first princi-

ples, such as a 3D volume model of the electromagnetic field

in neural tissue, increase confidence in the model; section 4.6

discussed how this gives the possibility of validating the Hodg-

kin-Huxley model and how upscaling techniques could be used

to derive, from the D volume model, a new model at the same

level of organization as the Hodgkin-Huxley model.

From the top-down perspective, doubt was raised whether a

D volume model of an entire network, such as those discussed

in section .., would really say anything, and, in particular,

that such very detailed network models lack measures of veri-

fiability. The approach suggested to the latter problem was to

verify the simulation tool on simpler problems, such as the test

problems discussed above.

In order to develop good test problems, there is a need for ex-

periments tailored towards simulation. It was pointed out that

there is a sociological, or structural problem in the field such

that experimentalists will not go very far in answering such

needs. On the other hand, there are now some experimental

laboratories which use modeling as a fundamental research

tool.

6.4 Model reproducibility

There was a consensus that published simulation results are, in

general, very hard to reproduce. Thus, the third major finding

of the workshop was:

Workshop participants reported on difficulties in repro-

ducing simulations from published articles.

There seem to be several common reasons for this:

· The model description is incomplete. It is often neces-

sary to contact the authors to get additional information

required to run the simulation.

· The model description differs from the model used to

produce the results. It occurs that last-minute modifica-

tions are made to the model, while the same modifica-

tions do not find their way into the methods section.

· The model description uses a formalism not supported

on the simulator used to reproduce the results. One

example of this is given in section ...

It may also happen that the simulations described in the ar-

ticle have been carried out on peculiar hardware and/or with

custom code, entailing substantial re-implementation efforts

when running the simulation with standard software on stan-

dard hardware.

Since reproducibility is one of the tenets of science, the work-

shop gives the following recommendation:

Develop publication guidelines for ensuring the reproduc-

ibility of simulations. INCF should develop guidelines for

publications with computer simulations and publish these on

the INCF homepage as a reference for authors. INCF should

also encourage publishers to implement policies for making

models available in conjunction with publication.

One topic discussed at the workshop was whether the concept

of reproducibility should include exact quantitative reproduc-

ibility, i.e. reproduction of not only the behavior of the model

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ 19 ]

but also the exact numerical values of simulation output. Such

a requirement is harsh, because it means not only that identical

algorithms must be used, but also that numerical operations

need to be performed in an identical order so that cancella-

tion effects, rounding errors, etc. do not cause noticeable dif-

ferences.

An intermediate requirement would be that every simulation

of the model with the same simulation software on the same

machine gives identical results. This kind of reproducibility is

important, for example during development and debugging.

Even this is far from trivial on a cluster, since this requires a

mechanism for seeding the pseudo random number generator

in each process. It may even be impossible to make the result

independent of the number of processes, because the uses of

random numbers occur in a different order depending on how

the model is distributed over processes.

6.5 Method development in computational

neuroscience

From a developers point of view, it was pointed out that the

strong commitment required to develop an application is not

sufficiently recognized. This is especially the case when mak-

ing the transition from research tool to public tool, because this

requires documentation, user support and high code quality, at

least in the case of an open source project. Workshop partici-

pants reported on the difficulty in persuading funding agencies

to fund simulator development. A quick round through the sim-

ulator developers present at the workshop showed that money

supporting the development consistently came from a project

not having simulator development as its main goal.

Large-scale simulation on parallel computers brings a new

degree of complexity into simulation methodology which in-

creases the need also for more theoretical method development.

Again, it seems as if such research falls between the chairs

with regard to funding--it is neither pure computer science

nor neuroscience. The workshop therefore gives the following

recommendation:

Encourage, support and fund work on method develop-

ment within computational neuroscience. INCF should in-

form funding agencies about the need for research on methods

in computational neuroscience, and on methods for large-scale

simulation in particular.

In a similar vein, workshop participants reported on the diffi-

culty in publishing simulation methods in peer-reviewed jour-

nals. One of the first peer reviewed journal articles on large-

scale simulation is Morrison et al. (00). The workshop gives

the following recommendation:

Encourage, support and fund the production of publica-

tions on concepts and techniques for large-scale simula-

tions. INCF should inform publishers about the need for

published work on methods within the field of large-scale sim-

ulations and computational neuroscience in general, and about

the current growth in this area.

6.6 A cyber infrastructure for

computational neuroscience?

Workshop participants recognized the need to collect models

into model repositories. This is important, not only for simu-

lator verification and benchmarking (section 6.2) but also to

aid development of large-scale network models. For example,

there is a need for a library of cell models to draw upon when

building network models.

Examples of existing model repositories are ModelDB (..1)

and the RIKEN J-Node portal/server (section ..1). Models in

these repositories can be coded in any specification language.

The only requirement is to supply a README file. In the fu-

ture, it may be desirable to develop model repositories con-

forming to a standard model description language (c.f. section

..).

6.6.1 The NRV project

The NRV project (Neuroinformation Research in Vision), at

the Japanese neuroinformatics node (http://www.neuroinf.jp/),

represents a mathematical modeling approach for understand-

ing the brain as a system through the integration of experimen-

tal data at multiple levels.

The current paradigm for information flow in science today

involves the use of many types of information technology, but

what is finally stored on the PC is scientific papers. The NRV

project aims to extend this paradigm by providing databases

with peer reviewed content which can be shared in a manner

similar to the sharing of papers. These databases contain:

· source code of models

· experimental data

· documents

The Visiome platform (Usui (00); Usui et al. (00); ..)

is a first example of such a database. Concepts from Visiome

have been used in development of the XooNIps content man-

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ 0 ]

agement system (..), which can be used as a base platform

for many types of database from small to large. XooNIps is

open-source software.

Other platforms at the Japanese neuroinformatics node in-

clude:

· The Neuron/Glia Platform, which integrates accumulat-

ing knowledge on the complex interactions between

neuron and glia, including the activities of associated

functional proteins.

· The Brain Machine Interface Platform (BMI PF), which

accumulates BMI (Brain Machine Interface)-related

experimental data, mathematical models, and tools

generated in neuroscience, computational theory, and

robotics.

· The Invertebrate Brain Platform (IVB PF), which

integrates experimental data, mathematical models,

and research tools relevant to the study of invertebrate

brains, neurons, and behavior.

· The Cerebellar Development Transcriptome Data-

base (CDT-DB), which provides spatio-temporal gene

expression profile information on the postnatal develop-

ment of mouse cerebellum. It seeks to reproduce the

genetic blueprint of cerebellar development.

Platforms for neuroimaging and integrative brain research are

currently under development.

Database platforms are integrated through the RIKEN J-Node

portal/server where the XooNIps system is the key feature for

inter-database communication.

The NRV project has as one goal to prepare for the arrival of

a 10 PFLOP next-generation supercomputer expected to be in-

stalled at RIKEN in 011.

It was observed by workshop participants that legislation in

some countries, for example Norway and Sweden, requires the

storage of research data for 10 years. It was suggested that this

kind of tool might be suitable for this task and such use would

entail many bonus benefits.

Another observation is that it would be motivating for individ-

ual researchers to contribute if publishing data in this kind of

digital database would entail an increased likelihood of fund-

ing.

6.6.2 GRID

GRID technology is a method of pooling computing resources,

such as computing, storage and network resources, so that they

can be shared on a demand basis. GRID computing was sug-

gested at the workshop as a possibility for sharing cluster re-

sources for large-scale simulations.

Markus Diesmann informed workshop participants about the

Black Forest Grid (Backofen et al. (00); ..1), which is a

collaborative effort of ten faculties of the Albert-Ludwigs Uni-

versity to establish joint grid computing facilities using the

framework and experience of the largest computing grid in

the world, the LHC (Large Hadron Collider) Computing Grid.

Such a framework is highly scalable and the software involved

has been successfully used at CERN to manage 1000 contrib-

uted CPUs and 000 TB of storage and to operate clusters of

more than 00 nodes. Using the grid infrastructure, the BFG

has founded the virtual organization (VO) CNS as a starting

point to explore the sharing of resources in Computational

Neuroscience (Potjans et al., 00).

Today, the GRID is not widely used for parallel process-

ing requiring low latency communication between the com-

pute nodes. However, this is just the constraint imposed by

large-scale simulations of neuronal networks. Thus, from the

Computational Neuroscience point of view, the grid would be

mainly used as a grid of high-performance clusters. Here, the

parallel simulation jobs are distributed across the grid, but each

job runs at a single site. Whether the global and local sched-

uling mechanisms are already suitable for this type of usage

needs to be explored.

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ 1 ]

Appendix A: Summary of

existing and planned tools

relevant to large-scale modeling

A.1 Model construction

A.1.1 L-Neuron

L-Neuron is a modeling tool for generating anatomically ac-

curate neuronal analogues based on Lindenmayer systems. It

is oriented towards single cell generation.

http://krasnow.gmu.edu/L-Neuron/

A.1.2 ArborVitae

ArborVitae (Senft and Ascoli, 1999) is software for the recon-

struction of networks by algorithmic amplification of morpho-

logical data. It can be used to generate large-scale, anatomi-

cally accurate networks

A.1.3 Neuroconstruct

neuroConstruct has been designed to simplify development of

complex networks of biologically realistic neurons, i.e. models

incorporating dendritic morphologies and realistic cell mem-

brane conductances. It is implemented in Java and generates

output for the NEURON and GENESIS simulators. It uses the

latest XML-based specifications for neuronal information ex-

change, including MorphML and ChannelML.

http://www.neuroconstruct.org/

A.1.4 NeuGen

NeuGen can generate realistic dendritic and axonal morphol-

ogy of neurons and networks in D and construct networks of

synaptically connected neurons in a cortical column. It uses

experimental data to extract anatomical "fingerprints" of cells

for synthetic generation.

http://neugen.uni-hd.de/

A.2 Simulators and environments

A.2.1 Parallel Neuron

The present standard distribution of NEURON, version .9,

supports parallel simulation of network models in which cells

on different processors are coupled by discrete logical spike

events.

http://www.neuron.yale.edu/neuron/docs/help/neuron/neuron/

classes/parcon.html#ParallelNetwork

A.2.2 Genesis 3

The GENESIS simulator is undergoing a major redevelopment

effort that will result in GENESIS version . Genesis has

a modular architecture based on plugins and has multiple in-

terfaces. This allows the use of multiple simulation engines,

such as modules from MOOSE or Neurospaces, to perform

the numerical calculations of the simulation. Models can be

specified in NeuroML, making it possible to easily exchange

models between simulators that use this standard. Advanced

Java-based network visualization programs can be used inter-

actively with GENESIS simulations. GENESIS can be di-

rectly interfaced to other simulators, in order to allow simula-

tions to interact. Simulations can be run from other modeling

environments, such as the Systems Biology Workbench or the

Modeler's Workspace. A WWW interface can allow remote

simulations to be run, using a web browser. Alternate parsers

can replace the GENESIS Script Language Interpreter to allow

the use of improved or more standard scripting languages, such

as Python. Estimated current development status is 1 yr before

modeling is possible for general users.

http://www.genesis-sim.org/GENESIS

http://moose-g.sourceforge.net

http://neurospaces.sourceforge.net

A.2.3 NEST

NEST is a simulation system for large networks of biologically

realistic (spiking) neurons. It is best suited for the simulation of

large networks of spiking point-neuron models where a realis-

tic number of synapses need to be represented. The internal dy-

namics of these models may be arbitrarily complex (or simple).

There is software support for heterogeneity in the neuron types

as well as in synaptic dynamics including short-term plasticity

and STDP. Neurons can interact by spikes constrained to the

simulation time grid or in continuous time. NEST uses a hy-

brid multi-threading and message-passing simulation scheme.

For more details see the corresponding section in Brette et al.

(00) as well as the following URLs.

http://www.scholarpedia.org/article/NEST

http://www.nest-initiative.org

A.2.4 NCS

NCS is a brain simulator which is optimized to model the hori-

zontally dispersed, vertically layered distribution of neurons

characteristic of the mammalian neocortex. The largest simula-

tions to-date have been on the order of a million single-com-

partment neurons and 1 trillion synapses.

http://brain.unr.edu/ncsDocs/

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ ]

A.2.5 CSIM and PCSIM

CSIM efficiently simulates medium sized networks of inte-

grate-and-fire neurons. It is extendable and has an interface to

Matlab. PCSIM provides a powerful environments for setting

up and simulating large-scale networks, performing data anal-

ysis and visualizing results. In PCSIM the Matlab interface has

been replaced with a Python interface.

http://www.lsm.tugraz.at/csim/

http://sourceforge.net/projects/pcsim

A.2.6 LENS

LENS is a simulation environment designed to support and in-

tegrate a wide range of brain modeling tasks across a variety of

computational platforms. In contrast to most neural modeling

environments, LENS supports arbitrary models and levels of

abstraction, from neural compartments, neurons, and micro-

circuits to brain structures and global brain processes. LENS

simulations are configured and initialized using a dynamically

extensible specification language in neuroscience domain

concepts and terms, which insulates model developers from

non-model-specific complexity. LENS is developed at Bio-

metaphorical Computing Research, IBM T.J. Watson Research

Center, Yorktown Heights, N.Y., U.S.A. For further informa-

tion, contact Charles Peck <cpeck@us.ibm.com>.

A.2.7 SPLIT

The SPLIT simulator (Hammarlund and Ekeberg, 1998) was

developed in the mid 90's with the aim of exploring how to

efficiently use the resources of various parallel computer ar-

chitectures for neural simulations. It is a C++ library which is

linked into the user program, which is serial, and can be linked

with a serial or parallel version of the library. Parallelism is

thus completely hidden from the user. SPLIT was recently op-

timized for the Blue Gene/L computer (Djurfeldt et al., 00)

and shows linear scaling behavior up to at least 09 proces-

sors. The largest simulation performed with SPLIT comprised

million neurons and 11 billion synapses. For further infor-

mation, contact Anders Lansner <ala@csc.kth.se>.

A.2.8 Systems Biology Workbench

The Systems Biology Workbench (SBW), is a software frame-

work that allows heterogeneous application components--writ-

ten in diverse programming languages and running on different

platforms--to communicate and use each others' capabilities

via a fast binary encoded-message system.

http://sbw.kgi.edu/research/sbwIntro.htm

A.3 Languages and language tools

A.3.1 NeuroML

NeuroML is an ongoing collaborative effort to support the use

of XML for software relating to neuroscience modeling. It is

not a single standard XML language - but rather a collection of

related XML projects for modeling different aspects and lev-

els of neural systems, from intracellular mechanisms and ion

channel kinetics to the dynamics of networks of reconstructed

neurons.

http://www.neuroml.org/

A.3.2 PyNN

The PyNN framework has been developed within the FACETS

project (Meier, 00) as a standard scripting language bind-

ing for neuronal network simulators. This abstracts differences

between simulators and provides a common way to specify

models and run simulations. The goal is that simulation scripts

in PyNN for simulator A will run on simulator B without modi-

fication. PyNN is based on the Python programming language.

The API has two parts, a low-level, procedural API, and a high-

level, object-oriented API which is intended to have a one-to-

one mapping with NeuroML.

http://pynn.gforge.inria.fr/

A.4 Visualization

A.4.1 DAVIS

DAVIS is a visualization system for cortical data and models.

http://vip.cs.utsa.edu/research/Davis/

A.5 Databases and Database tools

A.5.1 ModelDB

ModelDB provides an accessible location for storing and ef-

ficiently retrieving computational neuroscience models. Mod-

elDB is tightly coupled with NeuronDB (..). Models can

be coded in any language for any environment. Model code

can be viewed before downloading and browsers can be set to

auto-launch the models.

http://senselab.med.yale.edu/senselab/modeldb/

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ ]

A.5.2 NeuronDB

NeuronDB provides a dynamically searchable database of

three types of neuronal properties: voltage gated conductances,

neurotransmitter receptors, and neurotransmitter substances. It

contains tools that provide for integration of these properties in

a given type of neuron and compartment, and for comparison

of properties across different types of neurons and compart-

ments.

http://senselab.med.yale.edu/senselab/NeuronDB/

A.5.3 BioModels

http://www.ebi.ac.uk/biomodels/

A.5.4 Visiome

Visiome Platform is a digital research resource archive for vi-

sion science. The available resources include mathematical

models, experimental stimuli, experimental data, and analyti-

cal tools.

http://platform.visiome.neuroinf.jp/

A.5.5 XooNIps

XooNIps is an extension of the content management system

XOOPS for the needs of neuroinformatics. The XooNIps-based

database stores digital resources such as research articles, ex-

perimental data, mathematical models, and stimulations. It is a

modular base platform which can support many types of data-

base from small to large.

http://xoonips.sourceforge.jp/

A.6 GRID computing

A.6.1 Black Forest Grid initiative

The BFG is a collaborative effort of ten faculties of the Albert-

Ludwigs University to establish joint grid computing facilities

using the framework and experience of the largest computing

grid in the world, the LHC (Large Hadron Collider) Comput-

ing Grid, LHC.

http://www.bfg.uni-freiburg.de

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ ]

References

Aviel, Y., Mehring, C., Abeles, M., and Horn, D. (2003).

On embedding synfire chains in a balanced network.
Neural Computation, 15(6):13211340.

Backofen, R., Borrmann, H. G., Deck, W., Dedner,

A., Raedt, L. D., Desch, K., Diesmann, M., Gei-
er, M., Greiner, A., Hess, W. R., Honerkamp, J.,
Jankowski, S., Krossing, I., Liehr, A. W., Karwath,
A., Kloefkorn, R., Pesche, R., Potjans, T., Roettger,
M. C., L. Schmiedt-Thieme, G. S., Voss, B., Wiebelt,
B., Wienemann, P., and Winterer, V. H. (2006). A
bottom-up approach to grid-computing at a univer-
sity: the black-forest-grid initiative.

PIK - Praxis der

Informationsverarbeitung, 29:8187.

Bastian, P., Birken, K., Johannsen, K., Lang, S., Neuss,

N., Rentz-Reichert, H., and Wieners, C. (1997). Ug
- a flexible software toolbox for solving partial dif-
ferential equations.

Computing and Visualization in

Science, 1:2740.

Binzegger, T., Douglas, R. J., and Martin, K. A. C.

(2004). A quantitative map of the circuit of cat pri-
mary visual cortex.

J. Neurosci., 39:84418453.

Brette, R. (2006). Exact simulation of integrate-and-fire

models with synaptic conductances.

Neural Compu-

tation, 18(8):20042027.

Brette, R. and Gerstner, W. (2005). Adaptive exponen-

tial integrate-and-fire model as an effective descrip-
tion of neuronal activity.

J. Neurophysiol., 94:3637

3642.

Brette, R., Rudolph, M., Carnevale, T., Hines, M.,

Beeman, D., Bower, J. M., Diesmann, M., Mor-
rison, A., Goodman, P. H., Harris, F. C., Jr., Zirpe,
M., Natschläger, T., Pecevski, D., Ermentrout, B.,
Djurfeldt, M., Lansner, A., Rochel, O., Vieville,
T., Muller, E., Davison, A. P., Boustani, S. E., and
Destexhe, A. (2007). Simulation of networks of spik-
ing neurons: a review of tools and strategies.

J. Comp.

Neurosci. Submitted.

Brunel, N. (2000). Dynamics of sparsely connected net-

works of excitatory and inhibitory spiking neurons.
J. Comp. Neurosci., 8:183208.

Churchland, P. S. and Sejnowski, T. J. (1992).

The

Computational Brain. The MIT Press, Cambridge,
Massachusetts.

Denk, W. and Horstmann, H. (2004). Serial block-face

scanning electron microscopy to reconstruct three-
dimensional tissue nanostructure.

PLoS Biology,

2(11):19001909.

Destexhe, A., Rudolph, M., and Pare, D. (2003). The

high-conductance state of neocortical neurons in
vivo.

Nat. Rev. Neurosci., 4:739751.

Djurfeldt, M., Johansson, C., Örjan Ekeberg, Rehn, M.,

Lundqvist, M., and Lansner, A. (2005). Massively
parallel simulation of brain-scale neuronal network
models. Technical Report TRITA-NA-P0513, CSC,
KTH, Stockholm.

Eberhard, J., Attinger, S., and Wittum, G. (2004).

Coarse graining for upscaling of flow in hetero-
geneous porous media.

Multiscale Model. Simul.,

2(2):269301.

Einevoll, G. T., Pettersen, K. H., Devor, A., Ulbert, I.,

Halgren, E., and Dale, A. M. (2007). Laminar popu-
lation analysis: Estimating firing rates and evoked
synaptic activity from multielectrode recordings in
rat barrel cortex.

J. Neurophysiol. In press.

Fourcaud-Trocme, N., Hansel, D., van Vreeswijk, C.,

and Brunel, N. (2003). How spike generation mech-
anisms determine the neuronal response to fluctuat-
ing inputs.

J. Neurosci., 23:1162811640.

Haeusler, S. and Maass, W. (2007). A statistical analysis

of information-processing properties of lamina-spe-
cific cortical microcircuit models.

Cerebral Cortex,

17:149162.

Hammarlund, P. and Ekeberg, O. (1998). Large neu-

ral network simulations on multiple hardware plat-
forms.

J Comput Neurosci, 5(4):44359.

Izhikevich, E. M. (2003). Simple model of spiking neu-

rons.

IEEE Trans Neural Networks, 14:15691572.

Jolivet, R., Lewis, T. J., and Gerstner, W. (2004). Gener-

alized integrate-and-fire models of neuronal activity
approximate spike trains of a detailed model to a high
degree of accuracy.

J. Neurophysiol., 92:959976.

Keat, J., Reinagel, P., Reid, R. C., and Meister, M. (2001).

Predicting every spike: a model for the responses of
visual neurons.

Neuron, 30:803817.

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007

[ ]

Kennedy, H. (2005). The DAISY Project.

http://daisy.ini.

unizh.ch.

Kumar, A., Schrader, S., Aertsen, A., and Rotter, S.

(2007). The high-conductance state of cortical net-
works.

Neural Computation. In press.

Latham, P. E., Richmond, B. J., Nelson, O. G., and Ni-

renberg, S. (2000). Intrinsic dynamics in neuronal
networks. i. theory.

J. Neuophysiol., 83:808827.

Lundqvist, M., Rehn, M., Djurfeldt, M., and Lansner,

A. (2006). Attractor dynamics in a modular network
model of neocortex.

Network: Computation in Neu-

ral Systems, 17:253276.

MacGregor, R. J. and Oliver, R. M. (1974). A model for

repetitive firing in neurons.

Kybernetik, 16:5364.

Markram, H. and Peck, C. (2004). The Blue Brain Proj-

ect.

http://bluebrainproject.epfl.ch.

Marr, D. (1982).

Vision. Freeman, New York.

Mehring, C., Hehl, U., Kubo, M., Diesmann, M., and

Aertsen, A. (2003). Activity dynamics and propa-
gation of synchronous spiking in locally connected
random networks.

Biol. Cybern., 88(5):395408.

Meier, K. (2005). The FACETS Project.

http://facets.kip.

uni-heidelberg.de.

Morrison, A., Aertsen, A., and Diesmann, M. (2007a).

Spike-timing dependent plasticity in balanced ran-
dom networks.

Neural Computation. In press.

Morrison, A., Mehring, C., Geisel, T., Aertsen, A., and

Diesmann, M. (2005). Advancing the boundaries of
high-connectivity network simulation with distribut-
ed computing.

Neural Computation, 17:17761801.

Morrison, A., Straube, S., Plesser, H. E., and Diesmann,

M. (2007b). Exact subthreshold integration with con-
tinuous spike times in discrete-time neural network
simulations.

Neural Computation, 19(1):4779.

Paninski, L., Pillow, J. W., and Simoncelli, E. P. (2004).

Maximum likelihood estimation of a stochastic inte-
grate-and-fire neural encoding model.

Neural Com-

putation, 16:25332561.

Potjans, T., Wiebelt, B., and Diesmann, M. (2007). Grid

computing for computional neuroscience. In

10th

Tamagawa-Riken Dynamic Brain Forum, Hakuba,
Japan.

Richardson, M. J., Brunel, N., and Hakim, V. (2003).

From subthreshold to firing-rate resonance.

J. Neu-

rophysiol., 89:25382554.

Rudolph, M. and Destexhe, A. (2006). Analytical in-

tegrate-and-fire neuron models with conductance-
based dynamics for event-driven simulation strate-
gies.

Neural computation, 18(9):21462210.

Schutter, E. D. (1998). Dendritic voltage and calcium-

gated channels amplify the variability of postsynaptic
responses in a purkinje cell model.

Journal of Neuro-

physiology, 80:504519.

Senft, S. L. and Ascoli, G. A. (1999). Reconstruction of

brain networks by algorithmic amplification of mor-
phometry data.

Lecture Notes Computer Science,

1606:2533.

Sharp, A. A., O'Neil, M. B., Abbott, L. F., and Mard-

er, E. (1993). Dynamic clamp: Computer-generat-
ed conductances in real neurons.

J. Neurophysiol.,

69(3):992995.

Tetzlaff, T., Morrison, A., Geisel, T., and Diesmann, M.

(2004). Consequences of realistic network size on
the stability of embedded synfire chains.

Neurocom-

puting, 5860:117121.

Thomson, A. M., West, D. C., Wang, Y., and Bannister,

A. P. (2002). Synaptic connections and small circuits
involving excitatory and inhibitory neurons in layer
25 of adult rat and cat neocortex: Triple intracel-
lular recordings and biocytin labelling in vitro.

Cere-

bral Cortex, 12:939953.

Tuckwell, H. C. (1988).

Linear Cable Theory and Den-

dritic Structure, volume 1 of Introduction to Theo-

retical Neurobiology. Cambridge University Press,
Cambridge.

Usui, S. (2003). Visiome: neuroinformatics research in

vision project.

Neural Networks, 16:12931300.

Nature Precedings : doi:10.1038/npre.2007.262.1 : Posted 27 Jun 2007