=Paper=
{{Paper
|id=Vol-2022/paper25
|storemode=property
|title=
The EU's Human Brain Project (HBP) Flagship - Accelerating Brain Science Discovery and Collaboration
|pdfUrl=https://ceur-ws.org/Vol-2022/paper25.pdf
|volume=Vol-2022
|authors=Katrin Amunts
|dblpUrl=https://dblp.org/rec/conf/rcdl/Amunts17
}}
==
The EU's Human Brain Project (HBP) Flagship - Accelerating Brain Science Discovery and Collaboration
==
https://ceur-ws.org/Vol-2022/paper25.pdf
The EU's Human Brain Project (HBP) Flagship –
Accelerating brain science discovery and collaboration
(Extended Abstract)
© Katrin Amunts,
Jülich, Düsseldorf, Germany
Abstract. The human brain has a multi-level organisation and high complexity. New approaches are
necessary to decode the brain with its 86 billion nerve cells, which form extensive networks. Ultra-high
resolution models of the brain pose massive challenges in terms of data processing, visualisation and analysis.
The Human Brain Project creates a cutting-edge European infrastructure to enable cloud-based collaboration
among researchers from different disciplines around the world. The infrastructure includes development
platforms with databases, workflow systems, petabyte storage. New technologies, including neuromorphic
computing and robotics are being developed. Neuroscientists and ICT-specialists collaborate in a co-design
process, based on neuroscientific use cases, to develop such infrastructure, thus opening new perspective to
decode the human brain.
Keywords: data analytics, data management, future computing, neuroinformatics, neuroscience.
1 Introduction The HBP infrastructure will enable cloud-based
collaboration among researchers coming from
The human brain has a highly complex, multi-level different disciplines around the world. To achieve this,
organisation. New approaches are necessary to decode research platforms include databases, workflow
the brain with its 86 billion nerve cells, which form systems, petabyte storage, and supercomputers, to
complex networks – a challenge that is addressed by address the requirements of the different users.
the Human Brain Project (1). E.g., 3D Polarized Light
Imaging (2) elucidates the connectional architecture of 3 Project Structure
the brain at the level of axons, while keeping the
The core project involves 117 institutions and 500
topography of the whole organ; it results in data sets of
researchers from 19 countries. It is organized in 12
several Petabytes per brain, which should be actively
Subprojects. Neuroscience Subprojects include:
accessible while minimizing their transport. Thus,
Mouse Brain Organization, Human Brain
ultra-high-resolution models pose massive challenges
Organization, Systems and Cognitive Neuroscience
in terms of data processing, visualisation and analysis.
and Theoretical Neuroscience. Activities in these
High-resolution data obtained in post-mortem
Subprojects provide insights across all levels of brain
brains supplement information about structural and
organization: They will be more and more linked and
functional connectivity as obtained, e.g., by Magnetic
integrated into comprehensive models, and tested by
Resonance imaging, acquired from the living human
simulation.
brain, but with lower spatial resolution. By bringing
Big data analytics and simulation are highly
together different aspects of connectivity in one and
challenging and data intensive. Therefore, the Human
the same atlas, it is feasible to combine the advantages
Brain Project embraces ICT solutions. These include
of the different approaches, and to address the different
cloud-based collaboration and development platforms,
spatial (and temporal) scales.
with databases for metadata and provenance tracking,
2 The Human Brain Project as well as data analytics and compute services, right up
to leading-edge supercomputers, neuromorphic
The Human Brain Project is developing a new systems (4), and virtual robots (5). Research and
European research infrastructure. Part of it is the HBP development in these areas is organized in six
atlas. It integrates information from multiple levels of Subprojects, forming research platforms:
brain organisation including cellular architecture, Neuroinformatics, Simulation, High Performance
connectivity, molecular and genetic maps, as well Analytics and Computing, Medical Informatics,
results from neuroimaging and physiological studies. Neuromorphic Computing, and Robotics. In addition,
This atlas is a resource for empirical research, but also Management and Ethics & Society form two
modeling and simulation. To represent microscopical Subprojects.
data, it includes the Big Brain as one of its templates The main IT service infrastructures is represented
(3). by the Neuroinformatics Platform (NIP) and the High-
Performance Analytics and Computing Platform
Proceedings of the XIX International Conference (HPAC). Data from patients are being collected and
“Data Analytics and Management in Data Intensive analysed in the Medical Informatics Platform (6). The
Domains” (DAMDID/RCDL’2017), Moscow, Russia,
October 10–13, 2017
148
so-called COLLAB provides a connection between the https://education.humanbrainproject.eu/web/hbp-
Platforms, and a uniform entrance point. As a part of education-portal
the NIP it serves as the main collaboration
infrastructure. References
The development of the platforms is guided by co- [1] Amunts K, Ebell C, Muller J, Telefont M, Knoll A,
design projects where neuroscientific investigators and Lippert T.: The Human Brain Project: Creating a
IT specialists collaborate. After the Ramp-up phase of European research infrastructure to decode the
the HBP from 2013 to 2016, the platforms were human brain. Neuron 2016; 92 (3): 574–81 doi:
released in March 2016, can be accessed by test-users: 10.1016/j.neuron.2016.10.046
https://www.humanbrainproject.eu/en/. [2] Axer M, Amunts K, Grässel D, Palm C, Dammers
The NIP is conceptualized in such a way, that J, Axer H, Pietrzyk U, Zilles K.: A novel approach
researchers from other countries can share their data, to the human connectome: Ultra-high resolution
tools and expertise, to decode the human brain. This mapping of fiber tracts in the brain. Neuroimage
includes also collaboration with other brain initiatives 2011; 54: 1091–101. doi:
(7). 10.1016/j.neuroimage.2010.08.075
4 Supercomputing [3] Amunts K, Lepage C, Borgeat L, Mohlberg H,
Dickscheid T, Rousseau MÉ, Bludau S, Bazin PL,
To obtain the means needed to address the Lewis LB, Oros-Peusquens AM, Shah NJ, Lippert
incredible complexity of the brain, the neuroscience T, Zilles K, Evans AC.: BigBrain – an ultra-high
community will need to become a competitive player resolution 3D human brain model. Science 2013;
on high-end supercomputers and systems for big-data 340: 1472–5 doi: 10.1126/science.1235381
analytics. The HPAC aims to provide the Human Brain
[4] Friedmann S, Schemmel J, Grubel A, HArteil A,
Project and in extension the neuroscience community
Hock M, Meier K.: Demonstrating hyrid learning
with High Performance Computing systems geared
in a flexible neuromorphic hardware system. IEEE
towards the particular needs of neuroscientists.
Trans Biomed Circuits Syst 2017; 11 (1):128-142
Four European Tier 0 supercomputer centers
doi: 10.1109/TBCAS.2016.2579164. Epub 2016
currently are members of the consortium of the HBP:
Sep 9
the Barcelona Supercomputing Centre (BSC) in Spain,
the Consorzio Interuniversitario per le Applicazioni di [5] Falotico E, Vannucci L, Ambrosano A, Albanese
Supercalcolo per Università e Ricerca (CINECA) in U, Ulbrich S, Vasquez Tieck JC, Hinkel G, Kaiser
Italy, the Centro Svizzero di Calcolo Scientifico J, Peric I, Denninger O, Cauli N, Kirtay M,
(CSCS) in Lugano, Switzerland, and the Jülich Roennau A, Klinker G, Von Arnim A, Guyot L,
Supercomputing Centre (JSC) in Germany. Peppicelli D, Martínez-Cañada P, Ros E, Maier P,
At the JSC, two new pilot systems for an Weber S, Huber M, Plecher D, Röhrbein F, Deser
interactive supercomputer have been obtained as the S, Roitberg A, van der Smagt P, Dillman R, Levi
result of the “Pre-Commercial Procurement”: JULIA, P, Laschi C, Knoll AC, Gewaltig MO.: Connecting
developed by Cray, and JURON from IBM and artificial brains to robots in a comprehensive
NVIDIA. These systems are specifically designed for simulation framework: The Neurorobotics
data-intensive analytics and simulation applications in Platform. Front Neurorobot. 2017 Jan 25;11:2. doi:
the neurosciences. 10.3389/fnbot.2017.00002. eCollection 2017
[6] Frackowiak, R, Markram, H: The future of human
5 Teaching cerebral cartography: a novel approach. Philos
The convergence of neuroscience and ICT that the Trans R Soc Lond B Biol Sci. 2015 May 19;
Human Brain Project envisions in the long run depends 370(1668): 20140171. doi:
on training a new generation of young researchers to 10.1098/rstb.2014.0171
be fluent in both areas. To this end, the HBP Education [7] Vogelstein, J. T. & Neuro Cloud Consortium: To
Programme offers online courses, advanced HBP the Cloud! A Grassroots Proposal to Accelerate
Schools, an annual HBP Student Conference and Brain Science Discovery. Neuron. 2016 Nov
Young Researchers Events. 2;92(3):622-627. doi:
All educational offers are made available through 10.1016/j.neuron.2016.10.033
its website:
149