Project submission is closed
Registration is closed
Project submission information and description of break-out sessions
- Indicate which break-out groups are of interest in the online registration for: https://education.humanbrainproject.eu/web/young-researchers-event-2017/request-account-register
- Tell us about your projects, problems or what you would be interested to learn about the HBP Platforms and submit an abstract (like one of the example projects), no longer than 250 words (approx. 2000 chars.)
Project submission is closed.
Description of break-out groups and example projects
PROJECT 1 - Neuroinformatics Platform
guided by Michael von Papen (JUELICH)
Brief description of the platform: The Neuroinformatics Platform serves as HBP's search engine for data, brain atlases and knowledge about the brain. It also includes a set of standards, protocols and tools for building software applications. Users can search and collate neuroscience data and examine it by species, contributing laboratory, methodology, brain region, and data type. However, the Neuroinformatics Platform also provides toolkits for the analyses of data, which will be the topic of this project.
Example project: The goal of this project is to set up a basic 'collab' for the correlation analysis of parallel spike trains based on a Python notebook. We analyze the correlation structure between spike trains obtained from multielectrode arrays as this may reveal significant interaction and/or coding properties of different neuronal units. We create a Jupyter notebook within the HBP Collaboratory and upload data, documentation of the experiment and, of course, analysis routines and scripts for the calculation. We use the Electrophysiology Analysis Toolkit Elephant for the calculation of basic statistical neuronal properties. We calculate cross-correlation histograms and perform the significance tests. For this project you are highly encouraged to bring your own multielectrode data. Otherwise data is provided by the experts.
PROJECT 2 - Brain Simulation Platform
guided by Jean-Denis Courcol (EPFL)
Brief description of the platform: The Brain Simulation Platform (BSP) allows the user to reconstruct, analyze and simulate brain region models using a data-driven approach. While the data on any brain region is sparse, the BSP provides ways to fill the gaps making the process of simulating a brain region an achievable challenge.
In a continuous expansion, the BSP already contains tools, for example, to analyze experimental data such as single cell morphology reconstructions or recordings, optimize single cell and synapse models, reconstruct and simulate complex regions like the hippocampus.
- I am an experimentalist and I found an interesting result (e.g. on synaptic integration, network computation) which I cannot fully explain. I hope that a model of the single cell / network can help me to elucidate the phenomenon or guide me in future experiments. I have a series of data from my lab / literature which can be used to constrain such a model.
- I am a modeler and I am interested to test my hypothesis on the phenomenon X which occurs at the level of single cell / network (e.g. the synaptic integration in pyramidal cell is supra-linear and depends on current A, the network oscillation depends on the interplay between cell type B and C). Given the complexity of the phenomenon, a detailed model of single cell / network is the only way to go.
PROJECT 3 - High Performance Analytics and Computing Platform
guided by Lena Oden (JUELICH)
Brief description of the platform: The goal of the HPAC Platform is to build, integrate and operate the hardware and software components of the supercomputing, data and visualisation infrastructure required to:
- Run large-scale, data intensive, interactive multi-scale brain simulations up to the size of a full human brain.
- Manage the large amounts of data used and produced by the simulations, and
- Manage complex workflows comprising concurrent simulation, data analysis and visualisation workloads.
The HPC Platform currently provides supercomputers at four HPC centres for the HBP community: Barcelona Supercomputing Center (BSC), Cineca, Swiss National Supercomputing Centre (CSCS) and Jülich Supercomputing Centre (JSC).
More info at: https://hbp-hpc-platform.fz-juelich.de/
Example project: 3D-Polarized Light Imaging (3D-PLI) is a technique used to extract the orientation of nerve fibres at micrometer scale. Two different setups at the Institute of Neuroscience and Medicine (INM-1), Forschungszentrum Jülich are being used for the imaging.
A human brain is cut into about 2500 sections that are imaged using 3D-PLI microscopy, which results in about 1.3 micron pixel size and 80,000 x 100,000 pixels per section. The polarisation filters used in the imaging process are rotated 18 times so that 18 different images per section are available. The mentioned resolution and the 18 images result in a memory consumption of about 750 GB/section for the raw data, which is then transferred to the storage of the HPC systems at Jülich Supercomputing Centre (JSC), Forschungszentrum Jülich. The supercomputer JURECA is used to run the complex workflow that extracts the fibre orientations (3D vector per pixel) from the raw data. This analysis is not based on staining on the images but it uses the birefringence of the myelin sheaths that shield the axons of nerve fibres. Most tools in the workflow are developed at Forschungszentrum Jülich in the Fibre Architecture group led by Markus Axer and in close collaboration with Oliver Bücker and the Simulation Laboratory Neuroscience (both JSC). To speed up the workflow, it has been implemented using the UNICORE workflow engine (supported by André Giesler, JSC), which allows analysing a section in the order of hours instead of weeks. The goal of this use case project is now to integrate the existing UNICORE workflow into the infrastructure of the HBP.
PROJECT 4 - Medical Informatics Platform
guided by Mirco Nasuti (CHUV)
Brief description of the platform: The Medical Informatics Platform (MIP) is a data analysis system that facilitates the accurate diagnosis of brain disorders and their characterization, the identification of potential critical factors, and in general, the understanding of the brain and its functioning. The aim of the platform is to provide an interface between hospitals and research facilities where vast amounts of heterogeneous data is stored and processed in a regulated and harmonized way. Tools are provided for data mining, model estimation, statistical analysis, and classification, just to name a few.
Example project: Our goal is to verify the hypothesis (Xie et al., 2015) that changes in brain anatomy (specifically, in the atrophic regions) affect the alterations of intrinsic connectivity and result in disrupted cognitive performance of Alzheimer (AD) patients. Thus, we plan to identify brain regions with grey matter atrophy that is significantly associated with deficits in ADAS-cog 13 cognitive scores. Once the corresponding brain structures are localized, we intend to associate them to functional neuronal circuits that contribute to AD cognitive disruptions.
PROJECT 5 - Neuromorphic Computing Platform
guided by Petruţ Bogdan (UMAN)
Brief description of the platform: Neuromorphic computing relies on a few simplifying assumptions in order to achieve something quite unique; neuron models and learning rules are kept relatively simple in order to explore the behaviour of spiking neural networks. This translates to using point neuron models and learning rules which rely on local information to accelerate a spiking neural network simulation compared to traditional computing architectures, namely PCs, clusters or high performance computers. Two neuromorphic systems are available: BrainScaleS and SpiNNaker - an analogue implementation of spiking neurons and a multicore, software implementation of networks of these types of neurons.
Example project: I have a PyNN script or in some other simulation language, which represents a model of a basal ganglia. I want to use neuromorphic hardware to accelerate the simulation (a simulation which runs in 3 hours for 10 seconds of simulation could now run in 10 seconds on SpiNNaker or in 1 millisecond on BrainScaleS), but I also want to connect it to a small robot and observe its behaviour.
Counterexample project (this would not work for our systems):
I have a very high-fidelity model of a microcortical column which works when I simulate it on a supercomputer or my local cluster. It relies critically on using a neural model resembling Hodgkin and Huxley's.
PROJECT 6 - Neurorobotics Platform
guided by Axel von Arnim (Fortiss)
Brief description of the platform: The Neurorobotics Platform is bringing together experimenters and modellers to understand the neural basis of episodic memory as a faculty for spatiotemporal, multisensory integration, encoding and reconstruction. The techniques, including fMRI, electrophysiology, and imaging, are used for studying the neural mechanisms underlying the memory process in humans and rodents. The computational model and robotic implementation are being developed based on the experimental data.
Example project: One of the projects is to identify neural mechanisms integrating sensory modalities for episodic memory. We are studying the role of spatial cells in the hippocampus formation in this process.
We are currently developing a virtual reality environment where we can manipulate individual sensory inputs and investigate their effect on activity of spatial cells.
PROJECT 7 - NEST
guided by Dennis Terhorst (JUELICH)
More information on Project 7 will be published soon!