<style> .sticky-side-menu > a { display:block; margin-bottom: .2em; } .sticky-side-menu.stick { position:fixed; top: 50px; } @media screen and (max-width: 992px) { .sticky-side-menu.stick { display:block; position: static; } } </style> <script> var StickyMenu = function(containerId, anchorSelector) { var container = document.getElementById(containerId); var anchors = document.querySelectorAll(anchorSelector); var stickDistance = container.parentNode.offsetTop; var fixed = false; var resizeTmr = null; container.classList.add('sticky-side-menu'); window.addEventListener('resize', function() { if(resizeTmr) { clearTimeout(resizeTmr); } resizeTmr = setTimeout(setWidth, 10); }); var oAnchors = Array.from(anchors).map(function(anchor) { return { title: anchor.getAttribute('data-anchor'), target: anchor.getAttribute('id') } }); var generate = function() { for(var i = 0, j = oAnchors.length; i < j; i++) { var ref = oAnchors[i]; var a = document.createElement('a'); a.setAttribute('href', ['#', ref.target].join('')); a.appendChild(document.createTextNode(ref.title)); container.appendChild(a); } setWidth(); window.addEventListener('scroll', fixedMenu); } function fixedMenu () { var distance = stickDistance - window.pageYOffset; if(distance <= 0 && fixed === false) { container.classList.add('stick'); fixed = true; } else if (fixed === true && distance > 0) { container.classList.remove('stick'); fixed = false; } } function setWidth() { container.style.width = window.getComputedStyle(container.parentNode,null) .getPropertyValue("width"); } return { generate: generate } }; document.addEventListener('DOMContentLoaded', function(){ var sMenu = new StickyMenu('anchors-container', '[data-anchor]'); sMenu.generate(); }); </script>

Explanatory note

The original structure of the HBP grouped scientists by research focus in Subprojects as the primary operational structure.

For SGA3, which started in April 2020, the HBP took that trend a step further, using a small number of larger, cross-disciplinary Work Packages as its operational structure.

Co-design projects were multi-disciplinary and cross Subprojects. They were led by senior scientists from the HBP and were designed around collaboration, data gathering and simulation between the HBP’s Platforms.

Quick Menu






The Co-Design Projects (CDPs) have been chosen for their feasibility and potential value to different groups of target users. Working within and across Subprojects, each CDP aims to deliver a number of infrastructure product advances, which are a set of functionalities to achieve cross-Platform bridges and workflows in line with the objectives of the HBP. The CDPs are intended to drive these cross-Platform developments, involving tasks and investigators from both Neuroscience and Platform SPs, towards the realisation of an aligned Research Infrastructure. The HBP anticipates that the delivery of these products will generate new insights in neuroscience, medicine and computing.


Co-Design Project 1


Co-Design Project 1 - 
Development of the Whole Mouse Brain Model and the Related Mouse Brain Atlas

The first of the HBP Co-Design Projects, which involve multidisciplinary units from a number of Subprojects, consists of two components focused on research and engineering.

The first of these focuses on defining a number of scenarios (‘use cases') that represent some of the larger scientific questions that can be addressed using the HBP Platforms once the Co-Design Project has been completed. These research use cases will define which in silico preparations (e.g. whole brain models, slice models, etc.) the Platforms should provide, and how they can be measured and stimulated. The use cases also define the types of signals that can be measured in the in silico preparations, which will, in turn, dictate the granularity of the models (e.g. population/field models, point neuron networks, etc.) that should be used. The engineering component of the Co-Design Project will focus on developing Products based on the HBP Platforms that will make the use cases possible. For each of these Products, the researcher should be able to formulate the use case or related experiments in the Collaboratory, run the experiment in the Collaboratory, and access and analyse the results of the experiment in the Collaboratory or offline. A physical experimental platform will also be provided, to generate key missing mouse brain data to refine and validate the simulations.

CDP Science Leader: Francesco PAVONE

CDP Implementation Leader: Marc-Oliver GEWALTIG

Co-Design Project 2


Co-Design Project 2 - 
Mouse-Based Cellular Cortical and Sub-Cortical Microcircuit Models

CDP2 will integrate the activity of HBP platforms in order to refine brain modelling tools through their interactive use open to an enlarged scientific community. It aims to develop whole-brain scaffold models using detailed rodent cortical and sub-cortical microcircuits to integrate molecular / cellular properties into large-scale simulations. The models will exploit the base infrastructure (SP5, SP7) and the high-level infrastructure (SP6, SP10), and will be used for implementation in the Platforms (NIP, BSP, MIP, NRP) and to refine their tools through specific Use Cases. Fundamental simulations of brain activities and states in large-scale loops will be based on SP1/SP2 and community data and will address theoretical (SP3, SP4), medical (SP8) and technological (SP9, SP10) issues. 

CDP Science Leader: Egidio D'ANGELO

CDP Implementation Leader: Michele MIGLIORE

Co-Design Project 3


Co-Design Project 3 -
Multi-Level Human Brain Atlas

The overall aim of CDP3 is to develop a prototype of a comprehensive multimodal and multi-scale human brain atlas.

To support this, CDP3 will develop eight Products, which depend on components in several Tasks in different Subprojects. Most of the Products depend on a mixture of Tasks in several Subprojects, including SP2 (Strategic Human Brain Data), SP4 (Theoretical Neuroscience), SP5 (Neuroinformatics), SP7 (High Performance Analytics and Computing), SP8 (Medical Informatics) and SP12 (Ethics and Society). Since all of the CDPs focus on developing relevant and usable parts of the overall infrastructure that are highly scientifically relevant within the scope of SGA1, prioritisation and selection of the initial Atlas Products is required. The initial work of CDP3 in SGA1 will be the construction of a topographical representation of the human brain on different scales and considering different aspects, followed by a comprehensive semantic representation in future SGAs.

CDP Science Leader: Viktor JIRSA

CDP Implementation Leader: Timo DICKSCHEID

Co-Design Project 4


Co-Design Project 4 -
Visuo-Motor Integration

This CDP aims to develop a visuo-motor integration neural network model based on multi-level human neuroscience data. 

CDP4 aims to develop and implement multi-modal, neurobiologically realistic top-down models of visuo-motor integration that are implemented as running neuronal networks performing advanced object recognition and spatial object localisation to guide robotic motor control. Because of its central role, the project has focused first on eye-movement control, including the modelling of visual attention-for-action tasks that are required to derive parameter specifications for motor execution. CDP4 integrates ongoing work in SP2 and SP3 about realistic models of invariant object recognition and the role of attentional selection for both object recognition (ventral processing stream) and motor planning (dorsal processing stream). The empirically constrained models are implemented as NEST neural network models in collaboration with SP4, SP6 and SP7. The neural network models are implemented on the NRP in collaboration with SP10.

The visuo-motor architecture will be extended from eye movement control to reaching and grasping tasks allowing a common NEST network model to control arm and eye movements of a (virtual and real) robot in a closed-loop system able to perform non-trivial goal-directed behaviour. Furthermore, the visuo-motor integration architecture will be used to model unilateral spatial hemineglect resulting from right-hemispheric stroke in (predominantly) parietal cortex. Besides predicting behavioural neglect effects from different lesion sites from the running model, CDP4 aims to model also TMS treatment effects that are currently evaluated as a means to modulate the balance between the ipsi- and contralesional hemisphere in brain areas underlying saliency and spatial attention operations.

CDP Science Leader: Rainer GOEBEL

CDP Implementation Leader: Sacha VAN ALBEDA

Co-Design Project 5


Co-Design Project 5 -
Biological Deep Learning

This CDP aims to translate biological learning principles and algorithms to large-scale neuronal models simulating and implementing complex, multi-layer learning and related forms of cognition running on the simulation platforms of SPs 6, 7, 9, 10.

Deep learning represents a ground breaking progress in artificial intelligence that qualitatively changed the data-processing ability of computing devices. It remains unclear, however, how the brain could implement this type of learning, and whether the brain with its more complex neurons and signalling mechanisms found even better forms of implementing intelligent behaviour.

Biological forms of deep learning and related algorithms are studied in various subprojects in the HBP. At the same time, the neural simulation software and neuromorphic hardware developed in the HBP offer highly performant substrates for the implementation of such intelligent processing, in terms of both size and speed. CDP5 brings together these developments by linking cognition (SP3) and theory (SP4) to platform design and implementation (SP6, 7, 9, 10). The goal is to produce examples of brain-like intelligent computation in hardware that considerably goes beyond reproducing brain-like activity dynamics. Together with CDP4 (that focuses on visuo-motor transforms), CDP5 therefore adds a functional perspective to the data structures, models and platforms elaborated in the HBP.

CDP Science Leader: Mihai PETROVICI

CDP Implementation Leader: Walter SENN

Co-Design Project 6


Co-Design Project 6 -
Modelling Drug Discovery

The aim of this CDP is to develop new strategies for more effective drug treatments of major brain diseases such as Alzheimer’s, Schizophrenia, Epilepsy, Parkinson’s, glioblastoma and rare diseases using computational models. 

Innovative neuromedicine approaches require a detailed understanding of the molecular and systems-level organisation of the human brain, the causes and mechanisms of diseases, their progression, and the response to treatments. Because of the high level of complexity of the nervous system and the inter-subject variability in molecular brain organisation, behaviour and disease, addressing these issues for any neuropathology appears a daunting task. Indeed, for most neurodegenerative diseases, such as Alzheimer’s and Parkinson’s, there is currently no cure in spite of the very large investments from academia and industries. The discovery of new drugs against brain diseases is thus viewed as an ethical priority for ongoing neuroscience research.

Classical neuroactive drugs have been designed on the basis of their similarity-isosteric competitivity with compounds of natural origin. The allosteric interaction paradigm, instead, provides the distinction between the orthosteric ligands (binding to the endogenous neurotransmitter sites as agonists or antagonists), and ligands that mediate their effects by interacting at topographically distinct allosteric sites on the receptor. Molecular simulations, combined with experimental characterisation, will lead to the discovery of effective allosteric modulators and help to design new drugs with enhanced selectivity and reduced off-target effects.

CDP6 has the following objective: Design novel allosteric ligands with possible applications for diagnosis and therapeutic purposes. CDP6 will focus on developing molecular-based simulations tools to target allosteric sites in pharmaceutically relevant classes of biomolecules involved in neuropathologies. Libraries of candidate molecules will be provided for experimental tests within SP2 and SP8. The selected test-cases involve: Ligand- and voltage-gated ion channels, G protein Coupled receptors and PI3K.

CDP Science Leader: Jean-Pierre CHANGEUX

CDP Implementation Leaders: Paolo CARLONI and Zoe COURNIA

Co-Design Project 7


Co-Design Project 7 -
Hierarchical Planning during Navigation

This CDP aims to develop neurobiologically realistic models of hierarchical navigational planning in humans and rodents. It comprises an experimental stream and a computational stream. The experimental stream will produce novel rodent and human data, collected using benchmark paradigms for hierarchical planning in AI. The main outputs of the experimental stream will consist of novel datasets, software and atlases, with contributions to the HBP Subprojects SP1, SP2, SP3, SP5, and SP7. The computational stream will develop and implement neurobiologically realistic models of hierarchical navigational planning and integrate them in the HBP Platforms. For this, we will use a top-down modelling approach to translate abstract computational models of hierarchical planning into neural networks and, ultimately, embodied (robot) models. The main outputs of the computational stream in SGA2 will consist of novel, biologically realistic computational and robotic models of hierarchical planning, which will provide contributions to SP3, SP4, SP7, SP9, and SP10.

The fields of neuroscience and artificial intelligence (AI) share a goal of understanding how humans and other animals plan in complex natural environments. This CDP addresses this open problem by starting from the theory that, during navigation, rodents and humans plan using a hierarchical generative model of the task and its relevant states; and that this model is jointly implemented in the circuitry of the medial temporal lobe (MTL) and prefrontal cortex (PFC). This CDP will combine experimental (SP2, SP3) and computational (SP4, SP10) work streams to address the neural implementation of hierarchical spatial representations and of hierarchical inference, and develop biologically realistic computational and robotic models of equivalent abilities. The goal is to provide a detailed computational specification of an advanced cognitive ability – hierarchical planning – and integrate it within the HBP Platforms.

CDP Science Leader: Giovanni PEZZULO

CDP Implementation Leader: Hugo SPIERS

Co-Design Project 8


Co-Design Project 8 -
The Virtual Brain (TVB)

The aim of this CDP is to augment the existing simulation facilities of the HBP, which currently focus primarily on microscopic/mesoscopic simulation of spiking neurons and networks, by adding the large-scale (macroscopic) model implemented in TVB. TVB uses parallel implementations of neural mass and mean-field neural population models for efficient simulation of the whole brain. The CDP will provide code and interfaces that are based upon, and bridge to, the existing infrastructure provided by the HBP’s Brain Simulation Platform (BSP, SP6), Neuroinformatics Platform (NIP, SP5) and High-Performance Analytics and Computing Platform (HPAC SP7). A further aim is to provide external users with connectome extraction pipelines and preprocessed connectomes extracted from major public data repositories, along with educational aids to help users to get quickly acquainted with TVB tools. Lastly, connectivity features (e.g. density, directionality and laminar patterns) will be derived to augment human brain data sets.

The tools currently bundled on the HBP’s Brain Simulation Platform are mostly designed for microscopic and mesoscopic simulation of spiking neurons and networks. The Virtual Brain, on the other hand, simulates whole-brain activity on the basis of “neural mass” models that provide lumped descriptions of the activity of neural populations consisting of thousands of neurons. In this CDP, TVB’s large-scale simulation facilities will be added to the HBP’s code-base in order to provide software for parallelised, high-performance, multi-scale brain simulation. In addition to simulation functionalities, TVB includes tools for the preprocessing of MRI data, extraction of structural and functional connectomes and postprocessing of simulation results. A repository will be created, that will make available to external users fully preprocessed structural and functional connectomes extracted from major databases, such as the Human Connectome Project and the UK Biobank. Interactive educational tutorials will provide users with the necessary knowledge to use TVB tools. Multivariate analyses and machine learning techniques will be used to derive connectivity features based on information extracted from literature (such as the extensive, quantitative cytoarchitectonic data on the human cerebral cortex by von Economo and Koskinas).

CDP Science Leader: Petra RITTER

CDP Implementation Leader: Claus HILGETAG


updated Sep 2018