11-02-2013, 04:19 PM
Evaluation of the Blue Brain Project and Human Brain Project
[attachment=50550]
Introduction
The committee was requested to evaluate the Blue Brain Project (BBP) for its progress
since its last evaluation in early 2008, and for its role as the core for the Human Brain Project
(HBP) in its application for support by the ICT FET flagship initiative of the EC.
We organize this document accordingly, with sections for our review of the BBP followed
by the potential role of the BBP in the HBP. Given the complexity of the two projects, we indicate
the comprehensive organization and presentation by Professor Markram, and the evaluation by
our committee of each subproject of both the BBP and the HBP as well as the overall direction
and aims of both projects.
Executive Summary
Blue Brain Project (BBP)
The BBP aims at a quantitative simulation of the intrinsic connectivity of the cerebral
cortex with functional characterization of the neurons and microcircuits at an unprecedented
scale. Different classes of cortical neurons are identified by their morphological properties, by
their electrical properties in synaptically activating other neurons, as well as by their
pharmacological and gene expressing profile. This research program leads to a comprehensive
description of the wiring of a column of the cortex, which is being extended to neighboring
columns, and ultimately to the whole brain. Dr. Markram and his colleagues make a persuasive
case that this approach is critically important for understanding the neural basis of cortical
function, as well as for understanding the neural basis of neurological and psychiatric disorders.
The findings are based on research and data from cortical slices. These findings are
entered into databases which are linked to supercomputers which enable the information to be
retrieved for quantitative analysis and visualization. The research requires cutting-edge
supercomputing facilities to capture quantitatively the connectivity and physiological properties.
This large-scale and detailed approach to computational models of the cortex is in turn
influencing concepts underlying the development of the next generation of high performance
computing.
In summary, the research in the BBP, unique in its scope, is redefining the science and
technology of how to simulate the neural basis of brain function. The work in Switzerland is
closely linked to complementary research in other laboratories throughout Europe, and is the
central node of the Human Brain project. The opportunities for development of this work in the
next five years are great. Based on this evidence, we therefore recommend with the highest
enthusiasm the Blue Brain Project for an increased level of funding that will enable it to develop
considerably over the next five years.
Human Brain Project (HBP)
The Human Brain Project is a visionary multidisciplinary, multilevel, multi-laboratory
program that builds on the high-performance computer simulations of nerve cells and neuronal
microcircuits of the Blue Brain Project to enhance collaborative research in other critical areas of
technology applied to brain science.
In this innovative endeavor, the BBP provides a fundamental foundation for the HBP, in
that the ICT developed within the Blue Brain Project will be necessary for many of the
developments envisaged within the HBP to understand the brain, including the importance of
large scale computing to access the types of information that need to be stored, accessed and
used in models of brain function for many aspects of the HBP. A specific foundation provided by
the Blue Brain Project lies in the details of the connectivity of the cerebral cortex, and the ICT
facility that is being built for utilising that information, in for example the large scale simulations of
the cortex.
Evaluation of the BBP
This is a visionary project grounded in experimental analysis and computational synthesis
of neural properties and connectivity at the scale of a cortical column, in which a neuronal
network is generated from stored data and simulations run to investigate systematically the
performance of the network. As such the project is at the cutting edge in data intensive
computing, an emerging field within computational sciences that deals with the scientific data
explosion.
The project promises to have a profound impact on scientific computing in neuroscience.
But its potential goes far beyond the boundaries of neurobiology as it clearly has the potential to
transform biology as a whole by showing in a unique manner how the experimental-computational
integration should be done on the multiple scales necessary for understanding biological
processes including the brain. Furthermore, the specific requirements of the simulations of the
cerebral cortex will drive developments in high performance computing that will benefit other
areas of science.
Advancing Brain Science
As indicated above, the visit was highly organized and we were given a systematic
presentation of the entire project. We first briefly describe a sample of the scientific advances
presented in talks, lab visits and discussions that were evaluated during the visit.
Most of the work has focused on mini-columns in the somatosensory cortex (SA1) of the
rat, but the work has the potential to expand into other cortical areas. The rat is a good model in
which to develop the processes and build an initial inventory of all the neurons, their features,
physiological behaviors, and connections. These processes can then be repeated in any
(including higher) animals, including humans. Rodents are typical model organisms for biology
and most tools of gene expression etc are readily available. This enables a link between the
neuronal-level networks and systems biology.
An important scientific question that arises from the work of the BBP is the robustness of
the neuronal firing in the resting state of the local microcircuit. The computational work of the
group has demonstrated that these resting states are quite stable and invariant with respect to
perturbations of the neuronal connectivity. This is an empirical experimental result that raises new
questions of theoretical significance.
[attachment=50550]
Introduction
The committee was requested to evaluate the Blue Brain Project (BBP) for its progress
since its last evaluation in early 2008, and for its role as the core for the Human Brain Project
(HBP) in its application for support by the ICT FET flagship initiative of the EC.
We organize this document accordingly, with sections for our review of the BBP followed
by the potential role of the BBP in the HBP. Given the complexity of the two projects, we indicate
the comprehensive organization and presentation by Professor Markram, and the evaluation by
our committee of each subproject of both the BBP and the HBP as well as the overall direction
and aims of both projects.
Executive Summary
Blue Brain Project (BBP)
The BBP aims at a quantitative simulation of the intrinsic connectivity of the cerebral
cortex with functional characterization of the neurons and microcircuits at an unprecedented
scale. Different classes of cortical neurons are identified by their morphological properties, by
their electrical properties in synaptically activating other neurons, as well as by their
pharmacological and gene expressing profile. This research program leads to a comprehensive
description of the wiring of a column of the cortex, which is being extended to neighboring
columns, and ultimately to the whole brain. Dr. Markram and his colleagues make a persuasive
case that this approach is critically important for understanding the neural basis of cortical
function, as well as for understanding the neural basis of neurological and psychiatric disorders.
The findings are based on research and data from cortical slices. These findings are
entered into databases which are linked to supercomputers which enable the information to be
retrieved for quantitative analysis and visualization. The research requires cutting-edge
supercomputing facilities to capture quantitatively the connectivity and physiological properties.
This large-scale and detailed approach to computational models of the cortex is in turn
influencing concepts underlying the development of the next generation of high performance
computing.
In summary, the research in the BBP, unique in its scope, is redefining the science and
technology of how to simulate the neural basis of brain function. The work in Switzerland is
closely linked to complementary research in other laboratories throughout Europe, and is the
central node of the Human Brain project. The opportunities for development of this work in the
next five years are great. Based on this evidence, we therefore recommend with the highest
enthusiasm the Blue Brain Project for an increased level of funding that will enable it to develop
considerably over the next five years.
Human Brain Project (HBP)
The Human Brain Project is a visionary multidisciplinary, multilevel, multi-laboratory
program that builds on the high-performance computer simulations of nerve cells and neuronal
microcircuits of the Blue Brain Project to enhance collaborative research in other critical areas of
technology applied to brain science.
In this innovative endeavor, the BBP provides a fundamental foundation for the HBP, in
that the ICT developed within the Blue Brain Project will be necessary for many of the
developments envisaged within the HBP to understand the brain, including the importance of
large scale computing to access the types of information that need to be stored, accessed and
used in models of brain function for many aspects of the HBP. A specific foundation provided by
the Blue Brain Project lies in the details of the connectivity of the cerebral cortex, and the ICT
facility that is being built for utilising that information, in for example the large scale simulations of
the cortex.
Evaluation of the BBP
This is a visionary project grounded in experimental analysis and computational synthesis
of neural properties and connectivity at the scale of a cortical column, in which a neuronal
network is generated from stored data and simulations run to investigate systematically the
performance of the network. As such the project is at the cutting edge in data intensive
computing, an emerging field within computational sciences that deals with the scientific data
explosion.
The project promises to have a profound impact on scientific computing in neuroscience.
But its potential goes far beyond the boundaries of neurobiology as it clearly has the potential to
transform biology as a whole by showing in a unique manner how the experimental-computational
integration should be done on the multiple scales necessary for understanding biological
processes including the brain. Furthermore, the specific requirements of the simulations of the
cerebral cortex will drive developments in high performance computing that will benefit other
areas of science.
Advancing Brain Science
As indicated above, the visit was highly organized and we were given a systematic
presentation of the entire project. We first briefly describe a sample of the scientific advances
presented in talks, lab visits and discussions that were evaluated during the visit.
Most of the work has focused on mini-columns in the somatosensory cortex (SA1) of the
rat, but the work has the potential to expand into other cortical areas. The rat is a good model in
which to develop the processes and build an initial inventory of all the neurons, their features,
physiological behaviors, and connections. These processes can then be repeated in any
(including higher) animals, including humans. Rodents are typical model organisms for biology
and most tools of gene expression etc are readily available. This enables a link between the
neuronal-level networks and systems biology.
An important scientific question that arises from the work of the BBP is the robustness of
the neuronal firing in the resting state of the local microcircuit. The computational work of the
group has demonstrated that these resting states are quite stable and invariant with respect to
perturbations of the neuronal connectivity. This is an empirical experimental result that raises new
questions of theoretical significance.