4^th HBP Student Conference - Hands-on Workshop Day

The Human Brain Project has developed a new research infrastructure, EBRAINS, providing tools and services to the neuroscientific community. Researchers and students can share experimental and computational data, and the data is integrated to common 3D brain reference atlases to increase the comparability across modalities and scales. The EBRAINS data sharing service provides the necessary stewardship for sharing and integrating research data. By using organized and standardized metadata, we make the data easy to Find, Access, Interpret and Reuse in accordance with the FAIR principles (Wilkinson et al., Scientific Data 3:160018, 2016). Additionally, we use the anatomical location information available to map data to 3D brain reference atlases to achieve integration of neuroscience data regardless of their modality.  

In this workshop, we will present and demonstrate our workflows and curation services tailored for sharing heterogeneous neuroscience data through discussions and hands-on activities, and describe the steps from data generation to EBRAINS data publication. Participants will learn about data and metadata management and how to register neuroscience data to 3D brain reference atlases. We will also give an overview of tools and workflows available for integrating and analysing neuroscience data. For those interested we also offer individual guidance on how to best and most efficiently share own data on EBRAINS.

Ingrid Reiten is Msc in neuroscience and a researcher at the University of Oslo. With a background in biology, didactics and various modalities within neuroscience, she has developed an eagerness to couple these experiences to work towards a more collaborative research field with unified and more accessible research data. She does this in her current position as data curator in the Neuroinformatics platform of the Human Brain Project, collaborating across labs and research cultures to establish workflows for integration and sharing of heterogeneous neuroscience data. 

Ulrike Schlegel is Msc in Molecular Biology with specialisation in neuroscience and physiology, and currently works as a researcher at the University of Oslo. As part of the Human Brain Projects data curation team, she is working towards a more collaborative and transparent neuroscientific research culture. Her interdisciplinary background in Chemistry, Life Sciences and Molecular Biology gives her the suitable toolkit to understand the needs of researchers in natural sciences. 

The Neurorobotic Platform (NRP) is a tool developed in HBP SP10 to connect brains to robots in simulation. You can set-up experiments by declaring a robot, its actuators, sensors and environment and connect them to a brain simulation implemented in PyNN. Whatever your field, if you want to discuss ideas of an interdisciplinary experiment that can be rapidly prototyped and answer scientific questions that cannot be implemented in vivo, this subject is made for you.

Gabriel Urbain (MSc 2015) was born in Mons, Belgium, in 1990. After a short stay at the MIT Space System Labs, he graduated in 2015 with a Master in Electrical Engineering at the University of Mons (UMons), Belgium and a Master in Aerospace Engineering at Supaero, Toulouse, France. He is currently pursuing at PhD at the IDLab, Ghent University (UGent) -- imec, Belgium, and he is involved in the Neurorobotics branch of HBP. His current research includes locomotion on compliant robots, machine learning and morphological computation. In 2018, he was invited for three months at the Dynamic Legged System Lab from the Italian Institute of Technology (IIT).

Stefano Nardo is a software developer in The Biorobotics Institute of Scuola Superiore Sant'Anna. He received a Master's Degree in Computer Science at the University of Pisa in 2019 with a thesis on recurrent neural networks for sequence modeling.

With FENIX five European supercomputing centres, namely BSC (Spain), CEA (France), CINECA (Italy), CSCS (Switzerland) and JSC (Germany), have aligned their services to deliver as set of services requirements based primarily on neuroscience use-cases. The High Performance Analytics and Computing Platform develops and provides supercomputing, storage, visualization and simulation technology that can run on supercomputers and this FENIX infrastructure. Requesting FENIX resources is done with small proposal, which is internally reviewed. Student researchers are especially invited to apply for resources.
The first part of the session gives an introduction to the available resources and its basic usage in an hands-on session. In the second part the procedure for applying for these resources is explained. During the hands-on session students who want to applying for resources will be supported by experts from two HPC centers.

Experts: Wouter Klijn, Sandra Diaz, Alex Upton

Wouter Klijn completed an MSc in Artificial Intelligence from the University of Groningen, with a thesis on cortical micro-columns. He currently is a software architect in the Simlab Neuroscience part of the Jülich Supercomputer Centre, Forschungzentrum Jülich, with a focus on in AI, real-time big data streaming systems and complex HPC processing pipelines. He is responsible for science and use case management in the HBP, and ICEI/FENIX. He is also creating the science and software infrastructure architecture for the HBP.

This workshop will describe how to access the SpiNNaker platform, via both Jupyter Notebooks and the HBP Collaboratory.  It will then discuss how to write spiking neural networks using the PyNN language to be executed on SpiNNaker, and introduce the integration with the HBP Neurorobotics environment.  Participants will be given access to the Jupyter Notebook system from which they will be able to follow some lab examples, and then go on to create their own networks running on the platform, as well as create co-simulations with the robotics environment.

Structure:

1.      How to access SpiNNaker using Jupyter and the HBP Collaboratory

2.      How to use the NRP through the SpiNNaker Jupyter Service

3.      Running PyNN Simulations on SpiNNaker

4.      Run lab examples and write your own networks

Andrew Rowley is a Senior Research Software Engineer working at the University of Manchester in the UK, where he has been part of the SpiNNaker software development team since 2014, taking the role early in the Ramp-Up Phase of the Human Brain Project.  Andrew graduated from St. Andrews University with a BSc in Computer Science and Physics and a PhD in Computer Science studying Artificial Intelligence.  He then joined the University of Manchester as a Software Engineer in 2009, before working as a Senior Software Engineer at the National Centre for Text Mining also in the University.  His work is in leading the SpiNNaker software team in the development of all parts of the software, including the low-level light-weight “operating system”, the SpiNNaker C code that compiles into instruction code that runs on the machine, the Python and Java code that translates user scripts into executable networks, and the software that allows users to access the machine remotely.  He also likes to help researchers execute their neural network simulations on SpiNNaker, and has been involved in various tasks to this end, including the first verified execution of a cortical microcircuit on a neuromorphic platform.

Simulating a spiking network containing 1.000.000 neurons sounds like a daunting task? With the right high-performance computing hardware and the appropriate software, the NEST simulator, it is a piece of cake. Interested in a hands-on? Even if you just want to know how to use a high-performance computing system or how to quickly run a network simulation on your laptop, this will be useful for you. Now the interesting question: What would you do with both the computing power and the ability to simulate large-scale networks at hand?

Experts: Sandra Diaz Pier, Alper Yegenoglu, Alexander van Meegen

Sandra Diaz Pier was born in Mexico where she studied Electronic Systems Engineering and did a masters in Computer Sciences. Afterwards she moved to Canada where she obtained a second masters in Electrical Engineering. Today she works and is a PhD student at the Simulation Lab Neuroscience in the Jülich Research Centre in Germany. She is passionate about research in the fields of neuroscience and computation. Currently her research work involves efficient simulation of neural networks at different scales, structural plasticity, parameter space exploration techniques and analysis of the relationships between connectivity and function of the brain. Her work also involves helping neuroscientists port their applications and make efficient usage of the super computing infrastructure in Jülich.

Alper Yegenoglu received his Diploma (=M.Sc.) in 2013 in Computer Science at the Technical University of Dortmund where he worte his thesis on the extraktion and classification of salient regions in radiological data.
From 2013 to 2018 he worked as a software coordinator in the Institute of Neuroscience and Medicine (INM-6), Forschungszentrum Juelich, developing and cooridnating the Electrophysiology Analysis Toolkit (Elephant). In 2018 he started his PhD at the Simlab Neuroscience, Juelich Super Computing Center, Forschungszentrum Juelich. Currently he is working on machine learning techniques to understand how artificial algorithms can learn and how this differs from biological learning and how to bring those similar but different learning capabilities a little bit closer.

What can be done to prevent the misuse of neuroscience research? Should researchers and engineers worry about how the results of their work are used? How do we make sure that research findings and new technologies are used in responsible, ethical and socially beneficial ways? In this interactive workshop, we will address these questions (and many more!) with the aim of raising awareness of dual use and misuse issues among neuroscientists and engineers. The workshop will offer an overview of the fields of Dual Use, Misuse, and Responsible Research and Innovation, and will provide an opportunity to reflect on relevant developments in neuroscience and brain-related computer technologies. The workshop builds on the work of the HBP Dual Use working group.  

Tara Mahfoud is a Research Associate in the Department of Global Health and Social Medicine at King’s College London (KCL), and a Visiting Research Fellow in the Science, Technology and Society programme at the Harvard Kennedy School of Government. Tara is an anthropologist of science, technology and medicine interested in the cultural, social, political and clinical contexts and implications of developments in the neurosciences in Europe. Tara joined the HBP Ethics and Society sub-project in 2017 where she explores the potential future impacts of new knowledge and technologies, such as big data and machine learning, for neurology and psychiatry. At KCL, Tara also co-leads the Neuroscience and Society Network, an interdisciplinary forum for engagements across the social sciences, humanities and neurosciences. 

Inga Ulnicane has more than ten years of international and interdisciplinary research, teaching and engagement experience in the field of science, technology and innovation policy and governance. She has published on topics such as international research collaboration, European integration in science and technology and the Grand societal challenges concept as well as undertaken commissioned studies for European Parliament and European Commission. Currently she is a Research Fellow at De Montfort University (Leicester, UK). She is part of the HBP Ethics Support work package of the Ethics and Society subproject. She is Co-Chair of the Human Brain Project’s Dual Use working group and a task leader for Ethics Support Outreach and Dissemination. Her current research topics include governance of Artificial Intelligence and dual use.