Accelerate Your Neuroscience Research with EBRAINS

EBRAINS tutorials and users day - 12 March 2025 in Heidelberg (Germany)

The EBRAINS tutorials and users day is for two audiences: newcomers to the EBRAINS Research Infrastructure (introductions, overviews, beginner hands-on tutorials) and seasoned EBRAINS users for advanced tutorials and user discussion groups with developers.

• Quick demonstration of a sample set of "EBRAINS science tools in action" in the morning plenary talks (this part of the event can also be attended online free of charge)
• Followed by parallel hands-on tutorials: meeting participants can attend 3 consecutive sessions and choose for each from a set of in parallel tutorials. This part is onsite only for most tutorials.
  
  

Registration

• for attending on-site in Heidelberg (Germany) -- registration fee (coffee breaks, catering lunch and dinner at the meeting site, ...): 40 Euro for student attendants (Bachelor, Master, PhD students) / 80 Euro for everybody else.
  • Non-project members:
    • We strongly recommend to register with an EBRAINS account, as the account gives access to the services presented and an account will be needed to use the EBRAINS RI hands-on
      • get an EBRAINS account HERE (free of charge)
      • then register with EBRAINS account (with registration fee for the meeting)
    • It is also possible to register without an EBRAINS account (use this option primarily, if your email domain is not (yet) on the positive-list for immediate EBRAINS account creation)
  • EBRAINS2.0 project members please register here
    
    
• for attending on-line (via zoom) to (only) the plenary talks, i.e. the part until 11:00h -- free of charge:
  • register for on-line attendance with EBRAINS account (recommended)
  • or register for on-line attendance without EBRAINS account

Venue

European Institute for Neuromorphic Computing (EINC),
Im Neuenheimer Feld 225a,
D-69120 Heidelberg,
Germany

Airports

• Nearest airport: Frankfurt international airport (FRA)
  • Train Station in the airport itself: Frankfurt Flughafen Fernbahnhof (use bahn.de for time table information)
  • About an hour by car. Info: we often used the TLS shuttle service in the past. A shuttle for up to 7 people costs around 160 Euro single trip. This shuttle needs to be booked in advance.
• also relatively close: Stuttgart airport (STR) Train station in the airport: "Stuttgart Flughafen/Messe". (use bahn.de for time table information)

Train station

Heidelberg Main Station (Heidelberg Hauptbahnhof, use bahn.de for time table information)

From Heidelberg Main station to the institute:

• 25 min walking distance
• or 5 min by tram (Tram number 24, direction: Handschuhsheim Burgstraße. The tram departs from "Steig E" outside the station building. See the plan of Heidelberg Main station.
  Leave the tram at "Bunsengymnasium", then walk 3 min to the institute)

.

Time zone

The times in the agenda are in Central European Time CET (e.g. Europe, Berlin or Paris). Some other time zones: America, Australia, Europe, Japan, China, India)

(Only) the plenary talks can also be attended online. The tutorials are on site hands-on tutorials.

A high level introduction for researchers into the possibilities, tools and workflows offered by the EBRAINS Research Infrastructure today = for immediate use in own research projects

Briefly showing some of the tools / services / hardware systems or software environment, which are part of the EBRAINS research infrastructure in action -- "teasers" for the tutorials, starting after the coffee break.

(In parallel, 9:40h-11h: Facility Hub and Local Service Provider Meeting in the Seminar room R00.222 (ground floor)). This event is dedicated to the discussion and finalization of the Position Paper on the role of the Local Service Providers (including Facility Hubs), focused on accessory services to EBRAINS and their economic sustainability.

• Overview of EBRAINS Software Distribution (ESD) for a unified compute software environment across containerized environments (downloaded from EBRAINS OCI registry) for execution on the own machine, JupyterLab services and HPC resources
• Benefits of EBRAINS container images for use at home institutions
• Overview of pathways to acquire HPC resources

The EBRAINS Knowledge Graph:

• Documentation
• Search

Introductory tutorial

In this tutorial, we will learn how to search the EBRAINS Knowledge Graph and download data. We will introduce you to the openMINDS schemas and explain how metadata is organized in EBRAINS. Next, we will guide you through practical use cases, demonstrating how you can work with this data.

Participants will need an EBRAINS account, and at least a basic knowledge of Python programming.

Extended tutorial

This is a continuation of the introductory tutorial, covering further practical use cases.

Advanced tutorial

This tutorial will cover more advanced use of the EBRAINS Knowledge Graph, including more complex queries, uploading of metadata to the KG, and incorporating KG access into your own software.

Participants should already be familiar with the openMINDS metadata framework, either from attending the introductory tutorial or from reading the documentation, and should be able to program in either Python, Javascript, MATLAB or Java. Participants will need an EBRAINS account.

Tutors

Alix Bonard, Laura Morel, Peyman Najafi, Andrew Davison

We will present Arbor, a multicompartmental simulation library, that complements NEST, TVB, and nanoscale simulations, as well as offering interfaces to these tools. Arbor has been designed to leverage modern hardware, including GPUs, while delivering an intuitive interface to neuroscientists that is isolated from the concrete, low-level details. It has been shown to deliver performance up to the full scale of JUWELS booster. In this tutorial, we will show how to use Arbor starting from a simple cell to build a network of morphologically detailed cells. Participants will be given ample chance to interact the models.

Note: there are also two specific NEST related tutorials offered (see below):

• NESTML tutorial
• NEST astrocyte module tutorial

Tutorial offered: a beginners online hands-on experience on using NEST Desktop for teaching

NEST is a well known numeric spiking neural network simulator. NEST desktop brings the simulator to the browser for easy entry first-steps contact with NEST.

Pre Requires

• EBRAINS account for accessing Jupyter Hub at EBRAINS Lab is a must
• Knowledge on modeling, python, inference techniques would help

Agenda of the tutorial:

• Preparing the environment on EBRAINS Lab
• Working with tvb-library Python simulation module to obtain resting state …
• Inference exercise on Dynamical Causal Modeling of ERPs using Numpyro.

We will first introduce the building blocks of personalized virtual brain models with TVB through examples of simulating resting state activity (Lavanga et al. Neuroimage 2023, https://wiki.ebrains.eu/bin/view/Collabs/sga3-d1-2-showcase-1#HSimulationofresting-stateactivity) and epileptic seizures (Jirsa et al. Lancet Neurol. 2023), including virtual resections and stimulation. Related to the inference, we will provide a basic tutorial on Bayesian inference, illustrated with a dynamical causal modeling of event-related potentials observed in magneto/ electroencephalography (EEG/MEG) data introduced by David et al, Neuroimage 2006. Among probabilistic programming languages, we focused on NumPyro—a lightweight version of Pyro.ai that runs on the JAX framework. For automatic guide generation, this tutorial focuses on a self-tuning variant of Hamiltonian Monte Carlo, known as the NUTS sampler (Hofmman et al 2004). The taxonomy of algorithms and convergence diagnostics, developed in Task 3.3, is available here

Reference:

• Baldy, et. al 2024. Dynamic Causal Modeling in Probabilistic Programming Languages. bioRxiv. (see here)
• Lavanga, et al. 2023. The virtual aging brain: Causal inference supports interhemispheric dedifferentiation in healthy aging. NeuroImage. (see here)
• Jirsa et al. 2023. Personalised virtual brain models in epilepsy. Lancet Neurol. (see here)

Presenters

Spase Petkoski, Meysam Hashemi, Marmaduke Woodman, Lia Domide

In this tutorial, participants will be able to construct and simulate Spiking Neural Networks directly on the SpiNNaker hardware using the EBRAINS JupyterLab platform. They will learn how to program networks using the PyNN SNN language, and how the PyNN constructs work on the SpiNNaker platform. They will then get to try out these networks themselves and see the results from the simulations, as well as ask any other questions about SpiNNaker and how they might use it to explore SNNs in their own work.

The QUINT workflow takes you through steps to quantify and analyse labelled features within a known atlas space (2D rodent histological section images). The QUINT workflow comprises a suite of software designed to support atlas-based quantification. All the software have user interfaces, with no coding ability required. It generates object counts and percentage coverage per atlas-region, in addition to point clouds for visualising the features in 3D. For details see: https://www.ebrains.eu/brain-atlases/analysis/labelled-features-analysis

A prerequisite for using the workflow hands-in is to bring an own laptop with Windows installed and be prepared to install several software packages (it is rather quick to download and install them).

For downloading and installation instructions of the desktop versions, all instructions can be found on the website:

• as an overview
• For individual tools, you can find the “card” and the online manual from the quicknii card or the https://quicknii.readthedocs.io/en/latest/

Tutors

Sharon Yates and Maja Puchades

This tutorial will focus on the use of protein structural information to explore and predict molecular interactions and to estimate kinetic parameters (https://www.ebrains.eu/modelling-simulation-and-computing/simulation/molecular-and-subcellular-simulation). This information can be used to support the mathematical modeling of molecular signalling networks, particularly in studies of brain function and neurological disorders.

Following an overview of different approaches, the tutorial will focus on (1) τRAMD, a computationally efficient procedure that enables the computation of relative residence times (τ) or dissociation rates of protein-ligand complexes, and (2) SDA (Simulation of Diffusional Association), a Brownian dynamics simulation software package for the simulation of the diffusion of biomacromolecules in aqueous solution that can be used to compute bimolecular diffusional association rate constants.

Tutors: Riccardo Beccaria and Jonathan Teuffel

Tutorials offered: Introductory level and Advanced level.

Snudda (github link) builds networks of neurons with synaptic connectivity based on the morphologies of reconstructed neurons and touch detection. The detected synapses are pruned to match experimentally measured pairwise connectivity. The network connectivity can be exported, or directly simulated using NEURON.

Understanding the human brain requires access to experimental data that capture relevant aspects of brain organization across a broad range of scales and modalities, and typically originate from a plethora of resources. To make multimodal and multidimensional measures of brain organization accessible, they need to be integrated into a common reference framework and exposed via suitable software interfaces. This tutorial will introduce participants to siibra toolsuite, which provides access to a multilevel atlas of the human brain built from “big data”. The atlas integrates brain reference templates at different spatial scales, complementary parcellation maps, and a wide range of multimodal data features. It links macroanatomical concepts and their inter-subject variability with measurements of the microstructural composition and intrinsic variance of brain regions, using cytoarchitectonic maps as a reference, and integrating the BigBrain model as microscopic reference template. The tool suite includes a web-based 3D viewer (siibra-explorer) and a Python library (siibra-python) to support a broad range of neuroscientific use cases. It makes use of EBRAINS as a data sharing platform and cloud infrastructure and implements interfaces to other neuroscience resources. The focus of this tutorial will be on building reproducible workflows with BigBrain data using the siibra-python library.

Tutorials offered: Beginners level tutorial and a machine-learning methods related tutorial.

Beginner level tutorial (bio-inspired modelling): In this tutorial, participants will get individual remote access to the BrainScaleS-2 platform. They will learn how to program the system, explore its analog properties and configure complex neuron models. First, we will explore the dynamics of a leaky integrate-and-fire neuron on BrainScaleS-2, followed by the more complex adaptive exponential integrate-and-fire (AdEx) model. Then, we will introduce multi-compartment neuron models on BrainScaleS-2. Finally, we end our session by explaining how custom plasticity rules can be implemented on BrainScaleS-2. The session will be suitable for building basic knowledge about neuromorphic processors and analog computing as well as for kickstarting complex scientific research on BrainScaleS-2.

Tutor: Amani Atoui (UHEI)

Machine learning methods related tutorial: Recent advances in machine learning with neural networks require an increasing energy budget. Neuromorphic hardware platforms, such as the BrainScaleS-2 (BSS-2) system, aim to offer accelerated energy-efficient AI through analog computation by adopting event-based computing principles from biology. However, scalable training of spiking neural networks (SNNs) on an analog substrate remains challenging. We have developed a PyTorch-based machine learning-friendly framework for optimizing and partitioning SNNs on our analog hardware. In this hands-on tutorial, participants will learn how to train SNNs on BSS-2 using hxtorch.snn on a single neuron example and the YinYang dataset. We will also explain and showcase the process of partitioning larger-than-substrate networks onto multiple single-chips using the MNIST benchmark.

Tutor: Elias Arnold (UHEI)

We also offer full-length tutorials for these tools -- EBRAINS has more tools than short introduction talks fit into the morning...

Interactively build and optimize your own data-driven, biophysically detailed neuron model, via EBRAINS services and HPC facilities.

Tutorial offered: Introductory level

Biophysically detailed neural models help understand physiological mechanisms of individual cells and cell ensembles. In order to build a detailed model of a single neuron, the neuron morphology and the ion channel dynamics must be provided. Furthermore, a set of electrophysiological data is needed to fine tune the model parameters, if one wants to replicate the experimental observations.

In order to build and optimize biophysically detailed models of individual neurons, we have developed the Hodgkin-Huxley Neuron Builder (HHNB, hhnb.ebrains-italy.eu/hh-neuron-builder/) web application, where models, created via the NEURON simulation framework, are optimized against experimental recordings. Users can choose pre-existing models and data from the EBRAINS catalogues or provide their own. In addition, we have integrated the HHNB with the HippocampusHub (HH, hippocampushub.eu) where hippocampal morphologies, channel dynamics and neural signals can be selected from public repositories (e.g., neuromorpho.org, bluebrainnexus.io, modeldb.science) The focus of this turorial will be on the entire model building and optimization workflow, via the HH and the HHNB, from component selection to model simulation.

URL - (more details will be added to the agenda)

NESTML is a domain-specific language for neuron and synapse models.

The NESTML toolchain automatically generates high-performance code for these models, allowing them to be used in simulations of brain activity on several platforms, in particular the NEST Simulator running on HPC, but also the SpiNNaker neuromorphic platform. In this interactive tutorial, we will learn how to create a custom neuron model that exhibits dendritic action potentials, and combine it with a custom synaptic plasticity rule to create a network that learns and replays specific sequences of stimuli.

(The L2L tutorial had unfortunately to be cancelled)

In this introductory level tutorial we will introduce participants to analyzing electrophysiological data from simulation and experiment using the tools Neo (https://neuralensemble.org/neo) for representing and handling the data, and Elephant (https://python-elephant.org) for performing the actual analysis. We will cover using the basic data objects, loading a piece of realistic data and performing various data manipulations and analyses. For the tutorial, participants will be able to execute tutorials on their laptops with minimal installation effort (assuming working Python environment) or without any installation by using the EBRAINS Collaboratory online via browser.

Introductory level tutorial to the Brain Scaffold Builder (BSB), a component framework for network reconstrruction and simulations at multiple levels of detail

The tutorial will cover:

• Reconstruction of a network by BSB (https://github.com/dbbs-lab/bsb-core)
• Simulation of a point-neurons network by BSB-NEST (https://github.com/dbbs-lab/bsb-nest)
• Simulation of a multicompartmental-neurons network by BSB-NEURON (https://github.com/dbbs-lab/bsb-neuron)
• Development of own components in BSB

Cobrawap (Collaborative Brain Wave Analysis Pipeline) [1,2] is an open-source, modular and customizable data analysis tool developed in the context of HBP/EBRAINS, with the aim of enabling standardized quantitative descriptions of cortical wave dynamics observed in heterogeneous data sources, both experimental and simulated. The tool intercepts the increasing demand expressed by the Neuroscience community for reusability and reproducibility, offering a software framework suitable for collecting generalized implementations of established methods and algorithms, and for embracing innovative procedures. Inspired by FAIR principles and leveraging the latest findings in software engineering, Cobrawap is structured as a collection of modular Python3 building blocks that can be flexibly arranged along sequential stages, implementing data processing steps and analysis methods, directed by workflow managers (Snakemake or CWL). This “Getting started” tutorial [3] provides an introductory exercise to Cobrawap users, i.e. people interested in applying solutions already implemented in the software, for the exemplary scientific use-case of imaging recordings from mouse cortex under anesthesia. The value of this exercise goes beyond obtaining results from this specific use-case, since it allows also for “learning by doing” how the Cobrawap approach has been set up, and thus how it can be exploited in other scientific use-cases. Therefore, hints and clues will be given to people interested in developing new functions, i.e. expanding the pipeline to intercept their investigation goals and fit to their data types.

Requirements for attendees: a personal laptop for installing the software and storing test datasets and results.

References

[1] https://github.com/NeuralEnsemble/cobrawap, https://cobrawap.readthedocs.io
[2] Gutzen, et al. (2024) DOI 10.1016/j.crmeth.2023.100681
[3] https://github.com/APE-group/hands_on_cobrawap

Acknowledgments

Research co-funded by: European Union’s Horizon Europe Programme under Specific Grant Agreement No. 101147319 (EBRAINS 2.0); European Commission NextGeneration EU (PNRR EBRAINS-Italy MUR-CUP-B51E22000150006).

A three-parted demonstration session:

• the Medical Informatics Platform (MIP) demonstration will showcase the data catalog, including available federations and datasets. It will also demonstrate a machine learning algorithm from a clinician’s or researcher’s perspective—allowing users to select variables, configure parameters, and view intuitive visualizations via a user-friendly intuitive interface.
• and the Human Intracerebral EEG Platform (HIP) demo will showcase the HIP, highlighting its ability to facilitate secure data sharing, enable work in a private space, and support collaborative projects with restricted access to participating centers. It will also demonstrate its capacity to work in a remote desktop environment without leaving any fingerprint on personal computers or hospital infrastructure.
• The Platform for Human Imaging (PHI) demo will showcase PHI, an innovative platform developed from the HIP. Designed for storing, processing, and analyzing structural, functional, and diffusion MRI data, PHI is a fully remote system that provides a flexible working environment. Users can either develop code using a Bash terminal or take advantage of the user-friendly, intuitive applications integrated into the platform.

About the MIP

Presenter: Evita MAILLI (ATHENA RC) appr. 30 min plus 5 min for Q&A

The Medical Informatics Platform (MIP) is an advanced online collaborative platform designed for the scientific and medical community. It connects hospitals across a secure, decentralized network, allowing clinicians, clinical scientists, and researchers to identify disease patterns, improve diagnoses, and develop personalized treatments.

With MIP, users can:

• Access harmonized neurophysiological and medical data
• Explore research cohort datasets without transferring original clinical data
• Analyze patient information while ensuring privacy and security

This platform is the result of a multi-year, multidisciplinary effort involving computer scientists, clinical researchers, and medical professionals. Already deployed in 40+ hospitals across Europe, with 12 more installations in progress, MIP is transforming the way medical data is explored and utilized.

About the HIP

Presenter: Carolina CIUMAS (CHUV) appr. 30 min plus 5 min for Q&A

The Human Intracerebral EEG Platform HIP is a secure, open-source TRE (Trusted Research Environment) for collecting, storing, curating, sharing, and analysing multiscale human iEEG and imaging data internationally. It supports large-scale research projects and international collaborations by balancing data privacy, security, and access for accredited users. HIP covers iEEG-based research, including cognition, consciousness, connectomics, and related disorders, integrating advanced software and services for optimal data use. Adhering to FAIR principles, it respects ethics and data privacy regulations. HIP is planned to expand its focus to include multimodal neuroimaging, clinical, and biomarker data, as well as scalp-EEG and wearable device data, aiming to benchmark AI tools for seizure detection and prediction. It establishes a privacy-enabling data-sharing ecosystem with personal, collaborative, and public spaces, ensuring secure data management and compliance with legal and ethical standards

Outlook: CHORUS.HIP, the next version of the HIP
Presenter: Birgit SCHAFFHAUSER (CHUV) appr. 10 min

About the PHI

Presenters: Lorenzo PINI and Paolo Emilio Mazzon (UNIPD) appr. 20 minutes plus 5 min for Q&A

The Platform for Human Imaging (PHI) is the Padova incarnation of the HIP Platform developed at CHUV. It aims to provide the same level of experience when it comes to human data image processing. Within PHI, the Multimodal Analysis Toolbox for NeuroImaging (MATI v1.0) offers three main sections for selecting MRI processing and analysis workflows. It integrates both pre-existing imaging tools and in-house solutions that enhance functionality by enabling seamless software interaction and adding new features for processing, analysis, and quality control. Designed for flexibility, MATI v1.0 allows users to enter the workflow at any stage with pre-processed data, avoiding a rigid “black box” approach. Users can run only the necessary steps or execute the full analysis pipeline as needed.

Presenters: Lorenzo PINI and Paolo Emilio Mazzon (UNIPD)

EBRAINS tools for teaching brochure and overview poster

Info: The tool inroduction talks of this tutorial can also be attended online or on-site free of charge . Please find the info and registration for this option here. Special invitation letters for teachers to this session in English and in German.

Please find the speakers and any released slides and recordings here.

EBRAINS is a collaborative and digital European Research Infrastructure that provides access to a free and open database of neuroscience data, computational models, software tools and neuromorphic and high-performance computing systems for researchers, clinicians, scientists and students. Besides research purposes its tools are also a valuable resource for education and training.

In this session, aimed at educators from both university and high-school levels, we explore how EBRAINS tools can be effectively integrated into classroom settings to visualise and explain neuroscientific topics to students. Short presentations by tool experts will demonstrate practical examples of using these tools to enhance teaching and learning experiences. The session concludes with an open discussion on the opportunities, challenges, and requirements for adopting open science tools in neuroscience education.

Learning outcomes:

After attending this session, participants will

• understand how EBRAINS open science tools can be applied in neuroscience education.
• know how to integrate EBRAINS tools into classroom settings for visualising and explaining neuroscientific concepts.
• have discussed the opportunities and challenges of using open science tools in education.
• have gained ideas for incorporating EBRAINS tools into their own teaching practices.

Session Chairs:

Sandra Diaz, Forschungszentrum Jülich | Germany
Johannes Passecker, Medical University Innsbruck | Austria
Co-chairs of the EBRAINS Education Task Force

Speakers

• Timo Dickscheid (FZJ)
• Huifang Wang (AMU)
• Marmaduke Woodman (AMU)
• Sebastian Spreizer (UT)
• Luca Leonardo Bologna (CNR)
• Maja Puchades (UIO)
• Jakob Kaiser (UHEI)

Tool tutorial registrations will be collected in the registration step. We will then distribute the hands-on tutorials to the 3 parallel slots to maximize the possibility for attendants to visit their preferred tool tutorials.