A method to estimate the cellular composition of the mouse brain from heterogeneous datasets.

The mouse brain contains a rich diversity of inhibitory neuron types that have been characterized by their patterns of gene expression. However, it is still unclear how these cell types are distributed across the mouse brain. We developed a computational method to estimate the densities of different inhibitory neuron types across the mouse brain. Our method allows the unbiased integration of diverse and disparate datasets into one framework to predict inhibitory neuron densities for uncharted brain regions. We constrained our estimates based on previously computed brain-wide neuron densities, gene expression data from in situ hybridization image stacks together with a wide range of values reported in the literature. Using constrained optimization, we derived coherent estimates of cell densities for the different inhibitory neuron types. We estimate that 20.3% of all neurons in the mouse brain are inhibitory. Among all inhibitory neurons, 18% predominantly express parvalbumin (PV), 16% express somatostatin (SST), 3% express vasoactive intestinal peptide (VIP), and the remainder 63% belong to the residual GABAergic population. We find that our density estimations improve as more literature values are integrated. Our pipeline is extensible, allowing new cell types or data to be integrated as they become available. The data, algorithms, software, and results of our pipeline are publicly available and update the Blue Brain Cell Atlas. This work therefore leverages the research community to collectively converge on the numbers of each cell type in each brain region.

Optimum trajectory learning in musculoskeletal systems with model predictive control and deep reinforcement learning.

From the computational point of view, musculoskeletal control is the problem of controlling high degrees of freedom and dynamic multi-body system that is driven by redundant muscle units. A critical challenge in the control perspective of skeletal joints with antagonistic muscle pairs is finding methods robust to address this ill-posed nonlinear problem. To address this computational problem, we implemented a twofold optimization and learning framework to be specialized in addressing the redundancies in the muscle control . In the first part, we used model predictive control to obtain energy efficient skeletal trajectories to mimick human movements. The second part is to use deep reinforcement learning to obtain a sequence of stimulus to be given to muscles in order to obtain the skeletal trajectories with muscle control. We observed that the desired stimulus to muscles is only efficiently constructed by integrating the state and control input in a closed-loop setting as it resembles the proprioceptive integration in the spinal cord circuits. In this work, we showed how a variety of different reference trajectories can be obtained with optimal control and how these reference trajectories are mapped to the musculoskeletal control with deep reinforcement learning. Starting from the characteristics of human arm movement to obstacle avoidance experiment, our simulation results confirm the capabilities of our optimization and learning framework for a variety of dynamic movement trajectories. In summary, the proposed framework is offering a pipeline to complement the lack of experiments to record human motion-capture data as well as study the activation range of muscles to replicate the specific trajectory of interest. Using the trajectories from optimal control as a reference signal for reinforcement learning implementation has allowed us to acquire optimum and human-like behaviour of the musculoskeletal system which provides a framework to study human movement in-silico experiments. The present framework can also allow studying upper-arm rehabilitation with assistive robots given that one can use healthy subject movement recordings as reference to work on the control architecture of assistive robotics in order to compensate behavioural deficiencies. Hence, the framework opens to possibility of replicating or complementing labour-intensive, time-consuming and costly experiments with human subjects in the field of movement studies and digital twin of rehabilitation.

Feed-forward and noise-tolerant detection of feature homogeneity in spiking networks with a latency code.

In studies of the visual system as well as in computer vision, the focus is often on contrast edges. However, the primate visual system contains a large number of cells that are insensitive to spatial contrast and, instead, respond to uniform homogeneous illumination of their visual field. The purpose of this information remains unclear. Here, we propose a mechanism that detects feature homogeneity in visual areas, based on latency coding and spike time coincidence, in a purely feed-forward and therefore rapid manner. We demonstrate how homogeneity information can interact with information on contrast edges to potentially support rapid image segmentation. Furthermore, we analyze how neuronal crosstalk (noise) affects the mechanism's performance. We show that the detrimental effects of crosstalk can be partly mitigated through delayed feed-forward inhibition that shapes bi-phasic post-synaptic events. The delay of the feed-forward inhibition allows effectively controlling the size of the temporal integration window and, thereby, the coincidence threshold. The proposed model is based on single-spike latency codes in a purely feed-forward architecture that supports low-latency processing, making it an attractive scheme of computation in spiking neuronal networks where rapid responses and low spike counts are desired.

Experimental and Computational Study on Motor Control and Recovery After Stroke: Toward a Constructive Loop Between Experimental and Virtual Embodied Neuroscience.

Being able to replicate real experiments with computational simulations is a unique opportunity to refine and validate models with experimental data and redesign the experiments based on simulations. However, since it is technically demanding to model all components of an experiment, traditional approaches to modeling reduce the experimental setups as much as possible. In this study, our goal is to replicate all the relevant features of an experiment on motor control and motor rehabilitation after stroke. To this aim, we propose an approach that allows continuous integration of new experimental data into a computational modeling framework. First, results show that we could reproduce experimental object displacement with high accuracy via the simulated embodiment in the virtual world by feeding a spinal cord model with experimental registration of the cortical activity. Second, by using computational models of multiple granularities, our preliminary results show the possibility of simulating several features of the brain after stroke, from the local alteration in neuronal activity to long-range connectivity remodeling. Finally, strategies are proposed to merge the two pipelines. We further suggest that additional models could be integrated into the framework thanks to the versatility of the proposed approach, thus allowing many researchers to achieve continuously improved experimental design.

Corrigendum: A Cell Atlas for the Mouse Brain.

[This corrects the article DOI: 10.3389/fninf.2018.00084.].

Electrical Compartmentalization in Neurons.

The dendritic tree of neurons plays an important role in information processing in the brain. While it is thought that dendrites require independent subunits to perform most of their computations, it is still not understood how they compartmentalize into functional subunits. Here, we show how these subunits can be deduced from the properties of dendrites. We devised a formalism that links the dendritic arborization to an impedance-based tree graph and show how the topology of this graph reveals independent subunits. This analysis reveals that cooperativity between synapses decreases slowly with increasing electrical separation and thus that few independent subunits coexist. We nevertheless find that balanced inputs or shunting inhibition can modify this topology and increase the number and size of the subunits in a context-dependent manner. We also find that this dynamic recompartmentalization can enable branch-specific learning of stimulus features. Analysis of dendritic patch-clamp recording experiments confirmed our theoretical predictions.

A Cell Atlas for the Mouse Brain.

Despite vast numbers of studies of stained cells in the mouse brain, no current brain atlas provides region-by-region neuron counts. In fact, neuron numbers are only available for about 4% of brain of regions and estimates often vary by as much as 3-fold. Here we provide a first 3D cell atlas for the whole mouse brain, showing cell positions constructed algorithmically from whole brain Nissl and gene expression stains, and compared against values from the literature. The atlas provides the densities and positions of all excitatory and inhibitory neurons, astrocytes, oligodendrocytes, and microglia in each of the 737 brain regions defined in the AMBA. The atlas is dynamic, allowing comparison with previously reported numbers, addition of cell types, and improvement of estimates as new data is integrated. The atlas also provides insights into cellular organization only possible at this whole brain scale, and is publicly available.

Connecting Artificial Brains to Robots in a Comprehensive Simulation Framework: The Neurorobotics Platform.

Combined efforts in the fields of neuroscience, computer science, and biology allowed to design biologically realistic models of the brain based on spiking neural networks. For a proper validation of these models, an embodiment in a dynamic and rich sensory environment, where the model is exposed to a realistic sensory-motor task, is needed. Due to the complexity of these brain models that, at the current stage, cannot deal with real-time constraints, it is not possible to embed them into a real-world task. Rather, the embodiment has to be simulated as well. While adequate tools exist to simulate either complex neural networks or robots and their environments, there is so far no tool that allows to easily establish a communication between brain and body models. The Neurorobotics Platform is a new web-based environment that aims to fill this gap by offering scientists and technology developers a software infrastructure allowing them to connect brain models to detailed simulations of robot bodies and environments and to use the resulting neurorobotic systems for in silico experimentation. In order to simplify the workflow and reduce the level of the required programming skills, the platform provides editors for the specification of experimental sequences and conditions, environments, robots, and brain-body connectors. In addition to that, a variety of existing robots and environments are provided. This work presents the architecture of the first release of the Neurorobotics Platform developed in subproject 10 "Neurorobotics" of the Human Brain Project (HBP). At the current state, the Neurorobotics Platform allows researchers to design and run basic experiments in neurorobotics using simulated robots and simulated environments linked to simplified versions of brain models. We illustrate the capabilities of the platform with three example experiments: a Braitenberg task implemented on a mobile robot, a sensory-motor learning task based on a robotic controller, and a visual tracking embedding a retina model on the iCub humanoid robot. These use-cases allow to assess the applicability of the Neurorobotics Platform for robotic tasks as well as in neuroscientific experiments.

A Sparse Reformulation of the Green's Function Formalism Allows Efficient Simulations of Morphological Neuron Models.

We prove that when a class of partial differential equations, generalized from the cable equation, is defined on tree graphs and the inputs are restricted to a spatially discrete, well chosen set of points, the Green's function (GF) formalism can be rewritten to scale as O(n) with the number n of inputs locations, contrary to the previously reported O(n(2)) scaling. We show that the linear scaling can be combined with an expansion of the remaining kernels as sums of exponentials to allow efficient simulations of equations from the aforementioned class. We furthermore validate this simulation paradigm on models of nerve cells and explore its relation with more traditional finite difference approaches. Situations in which a gain in computational performance is expected are discussed.

Dynamics of self-sustained asynchronous-irregular activity in random networks of spiking neurons with strong synapses.

Random networks of integrate-and-fire neurons with strong current-based synapses can, unlike previously believed, assume stable states of sustained asynchronous and irregular firing, even without external random background or pacemaker neurons. We analyze the mechanisms underlying the emergence, lifetime and irregularity of such self-sustained activity states. We first demonstrate how the competition between the mean and the variance of the synaptic input leads to a non-monotonic firing-rate transfer in the network. Thus, by increasing the synaptic coupling strength, the system can become bistable: In addition to the quiescent state, a second stable fixed-point at moderate firing rates can emerge by a saddle-node bifurcation. Inherently generated fluctuations of the population firing rate around this non-trivial fixed-point can trigger transitions into the quiescent state. Hence, the trade-off between the magnitude of the population-rate fluctuations and the size of the basin of attraction of the non-trivial rate fixed-point determines the onset and the lifetime of self-sustained activity states. During self-sustained activity, individual neuronal activity is moreover highly irregular, switching between long periods of low firing rate to short burst-like states. We show that this is an effect of the strong synaptic weights and the finite time constant of synaptic and neuronal integration, and can actually serve to stabilize the self-sustained state.

Current practice in software development for computational neuroscience and how to improve it.

Almost all research work in computational neuroscience involves software. As researchers try to understand ever more complex systems, there is a continual need for software with new capabilities. Because of the wide range of questions being investigated, new software is often developed rapidly by individuals or small groups. In these cases, it can be hard to demonstrate that the software gives the right results. Software developers are often open about the code they produce and willing to share it, but there is little appreciation among potential users of the great diversity of software development practices and end results, and how this affects the suitability of software tools for use in research projects. To help clarify these issues, we have reviewed a range of software tools and asked how the culture and practice of software development affects their validity and trustworthiness. We identified four key questions that can be used to categorize software projects and correlate them with the type of product that results. The first question addresses what is being produced. The other three concern why, how, and by whom the work is done. The answers to these questions show strong correlations with the nature of the software being produced, and its suitability for particular purposes. Based on our findings, we suggest ways in which current software development practice in computational neuroscience can be improved and propose checklists to help developers, reviewers, and scientists to assess the quality of software and whether particular pieces of software are ready for use in research.

Does spike-timing-dependent synaptic plasticity couple or decouple neurons firing in synchrony?

Spike synchronization is thought to have a constructive role for feature integration, attention, associative learning, and the formation of bidirectionally connected Hebbian cell assemblies. By contrast, theoretical studies on spike-timing-dependent plasticity (STDP) report an inherently decoupling influence of spike synchronization on synaptic connections of coactivated neurons. For example, bidirectional synaptic connections as found in cortical areas could be reproduced only by assuming realistic models of STDP and rate coding. We resolve this conflict by theoretical analysis and simulation of various simple and realistic STDP models that provide a more complete characterization of conditions when STDP leads to either coupling or decoupling of neurons firing in synchrony. In particular, we show that STDP consistently couples synchronized neurons if key model parameters are matched to physiological data: First, synaptic potentiation must be significantly stronger than synaptic depression for small (positive or negative) time lags between presynaptic and postsynaptic spikes. Second, spike synchronization must be sufficiently imprecise, for example, within a time window of 5-10 ms instead of 1 ms. Third, axonal propagation delays should not be much larger than dendritic delays. Under these assumptions synchronized neurons will be strongly coupled leading to a dominance of bidirectional synaptic connections even for simple STDP models and low mean firing rates at the level of spontaneous activity.

Cortext: a columnar model of bottom-up and top-down processing in the neocortex.

Experimental data suggests that a first hypothesis about the content of a complex visual scene is available as early as 150 ms after stimulus presentation. Other evidence suggests that recognition in the visual cortex of mammals is a bidirectional, often top-down driven process. Here, we present a spiking neural network model that demonstrates how the cortex can use both strategies: Faced with a new stimulus, the cortex first tries to catch the gist of the scene. The gist is then fed back as global hypothesis to influence and redirect further bottom-up processing. We propose that these two modes of processing are carried out in different layers of the cortex. A cortical column may, thus, be primarily defined by the specific connectivity that links neurons in different layers into a functional circuit. Given an input, our model generates an initial hypothesis after only a few milliseconds. The first wave of action potentials traveling up the hierarchy activates representations of features and feature combinations. In most cases, the correct feature representation is activated strongest and precedes all other candidates with millisecond precision. Thus, our model codes the reliability of a response in the relative latency of spikes. In the subsequent refinement stage where high-level activity modulates lower stages, this activation dominance is propagated back, influencing its own afferent activity to establish a unique decision. Thus, top-down influence de-activates representations that have contributed to the initial hypothesis about the current stimulus, comparable to predictive coding. Features that do not match the top-down prediction trigger an error signal that can be the basis for learning new representations.

Towards reproducible descriptions of neuronal network models.

Progress in science depends on the effective exchange of ideas among scientists. New ideas can be assessed and criticized in a meaningful manner only if they are formulated precisely. This applies to simulation studies as well as to experiments and theories. But after more than 50 years of neuronal network simulations, we still lack a clear and common understanding of the role of computational models in neuroscience as well as established practices for describing network models in publications. This hinders the critical evaluation of network models as well as their re-use. We analyze here 14 research papers proposing neuronal network models of different complexity and find widely varying approaches to model descriptions, with regard to both the means of description and the ordering and placement of material. We further observe great variation in the graphical representation of networks and the notation used in equations. Based on our observations, we propose a good model description practice, composed of guidelines for the organization of publications, a checklist for model descriptions, templates for tables presenting model structure, and guidelines for diagrams of networks. The main purpose of this good practice is to trigger a debate about the communication of neuronal network models in a manner comprehensible to humans, as opposed to machine-readable model description languages. We believe that the good model description practice proposed here, together with a number of other recent initiatives on data-, model-, and software-sharing, may lead to a deeper and more fruitful exchange of ideas among computational neuroscientists in years to come. We further hope that work on standardized ways of describing--and thinking about--complex neuronal networks will lead the scientific community to a clearer understanding of high-level concepts in network dynamics, and will thus lead to deeper insights into the function of the brain.

Structural plasticity controlled by calcium based correlation detection. helias@bccn.uni-freiburg.de.

Hebbian learning in cortical networks during development and adulthood relies on the presence of a mechanism to detect correlation between the presynaptic and the postsynaptic spiking activity. Recently, the calcium concentration in spines was experimentally shown to be a correlation sensitive signal with the necessary properties: it is confined to the spine volume, it depends on the relative timing of pre- and postsynaptic action potentials, and it is independent of the spine's location along the dendrite. NMDA receptors are a candidate mediator for the correlation dependent calcium signal. Here, we present a quantitative model of correlation detection in synapses based on the calcium influx through NMDA receptors under realistic conditions of irregular pre- and postsynaptic spiking activity with pairwise correlation. Our analytical framework captures the interaction of the learning rule and the correlation dynamics of the neurons. We find that a simple thresholding mechanism can act as a sensitive and reliable correlation detector at physiological firing rates. Furthermore, the mechanism is sensitive to correlation among afferent synapses by cooperation and competition. In our model this mechanism controls synapse formation and elimination. We explain how synapse elimination leads to firing rate homeostasis and show that the connectivity structure is shaped by the correlations between neighboring inputs.

PyNEST: A Convenient Interface to the NEST Simulator.

The neural simulation tool NEST (http://www.nest-initiative.org) is a simulator for heterogeneous networks of point neurons or neurons with a small number of compartments. It aims at simulations of large neural systems with more than 10(4) neurons and 10(7) to 10(9) synapses. NEST is implemented in C++ and can be used on a large range of architectures from single-core laptops over multi-core desktop computers to super-computers with thousands of processor cores. Python (http://www.python.org) is a modern programming language that has recently received considerable attention in Computational Neuroscience. Python is easy to learn and has many extension modules for scientific computing (e.g. http://www.scipy.org). In this contribution we describe PyNEST, the new user interface to NEST. PyNEST combines NEST's efficient simulation kernel with the simplicity and flexibility of Python. Compared to NEST's native simulation language SLI, PyNEST makes it easier to set up simulations, generate stimuli, and analyze simulation results. We describe how PyNEST connects NEST and Python and how it is implemented. With a number of examples, we illustrate how it is used.

Interoperability of neuroscience modeling software: current status and future directions.

Neuroscience increasingly uses computational models to assist in the exploration and interpretation of complex phenomena. As a result, considerable effort is invested in the development of software tools and technologies for numerical simulations and for the creation and publication of models. The diversity of related tools leads to the duplication of effort and hinders model reuse. Development practices and technologies that support interoperability between software systems therefore play an important role in making the modeling process more efficient and in ensuring that published models can be reliably and easily reused. Various forms of interoperability are possible including the development of portable model description standards, the adoption of common simulation languages or the use of standardized middleware. Each of these approaches finds applications within the broad range of current modeling activity. However more effort is required in many areas to enable new scientific questions to be addressed. Here we present the conclusions of the "Neuro-IT Interoperability of Simulators" workshop, held at the 11th computational neuroscience meeting in Edinburgh ( July 19-20 2006; http://www.cnsorg.org ). We assess the current state of interoperability of neural simulation software and explore the future directions that will enable the field to advance.