Brain simulation as a cloud service: The Virtual Brain on EBRAINS.

The Virtual Brain (TVB) is now available as open-source services on the cloud research platform EBRAINS (ebrains.eu). It offers software for constructing, simulating and analysing brain network models including the TVB simulator; magnetic resonance imaging (MRI) processing pipelines to extract structural and functional brain networks; combined simulation of large-scale brain networks with small-scale spiking networks; automatic conversion of user-specified model equations into fast simulation code; simulation-ready brain models of patients and healthy volunteers; Bayesian parameter optimization in epilepsy patient models; data and software for mouse brain simulation; and extensive educational material. TVB cloud services facilitate reproducible online collaboration and discovery of data assets, models, and software embedded in scalable and secure workflows, a precondition for research on large cohort data sets, better generalizability, and clinical translation.

On the influence of prior information evaluated by fully Bayesian criteria in a personalized whole-brain model of epilepsy spread.

Individualized anatomical information has been used as prior knowledge in Bayesian inference paradigms of whole-brain network models. However, the actual sensitivity to such personalized information in priors is still unknown. In this study, we introduce the use of fully Bayesian information criteria and leave-one-out cross-validation technique on the subject-specific information to assess different epileptogenicity hypotheses regarding the location of pathological brain areas based on a priori knowledge from dynamical system properties. The Bayesian Virtual Epileptic Patient (BVEP) model, which relies on the fusion of structural data of individuals, a generative model of epileptiform discharges, and a self-tuning Monte Carlo sampling algorithm, is used to infer the spatial map of epileptogenicity across different brain areas. Our results indicate that measuring the out-of-sample prediction accuracy of the BVEP model with informative priors enables reliable and efficient evaluation of potential hypotheses regarding the degree of epileptogenicity across different brain regions. In contrast, while using uninformative priors, the information criteria are unable to provide strong evidence about the epileptogenicity of brain areas. We also show that the fully Bayesian criteria correctly assess different hypotheses about both structural and functional components of whole-brain models that differ across individuals. The fully Bayesian information-theory based approach used in this study suggests a patient-specific strategy for epileptogenicity hypothesis testing in generative brain network models of epilepsy to improve surgical outcomes.

Domain and interdomain energetics underlying gating in Shaker-type Kv channels.

To understand gating events with a time-base many orders-of-magnitude slower than that of atomic motion in voltage-gated ion channels such as the Shaker-type KV channels, a multiscale physical model is constructed from the experimentally well-characterized voltage-sensor (VS) domains coupled to a hydrophobic gate. The four VS domains are described by a continuum electrostatic model under voltage-clamp conditions, the control of ion flow by the gate domain is described by a vapor-lock mechanism, and the simple coupling principle is informed by known experimental results and trial-and-error. The configurational energy computed for each element is used to produce a total Hamiltonian that is a function of applied voltage, VS positions, and gate radius. We compute statistical-mechanical expectation values of macroscopic laboratory observables. This approach stands in contrast with molecular-dynamic models which are challenged by increasing scale, and kinetic models which assume a probability distribution rather than derive it from the underlying physics. This generic model predicts well the Shaker charge/voltage and conductance/voltage relations; the tight constraints underlying these results allow us to quantitatively assess the underlying physical mechanisms. The total electrical work picked up by the VS domains is an order-of-magnitude larger than the work required to actuate the gate itself, suggesting an energetic basis for the evolutionary flexibility of the voltage-gating mechanism. The cooperative slide-and-interlock behavior of the VS domains described by the VS-gate coupling relation leads to the experimentally observed bistable gating. This engineering approach should prove useful in the investigation of various elements underlying gating characteristics and degraded behavior due to mutation.

The sliding-helix voltage sensor: mesoscale views of a robust structure-function relationship.

The voltage sensor (VS) domain of voltage-gated ion channels underlies the electrical excitability of living cells. We simulate a mesoscale model of the VS domain to determine the functional consequences of some of its physical elements. Our mesoscale model is based on VS charges, linear dielectrics, and whole-body motion, applied to an S4 "sliding helix." The electrostatics under voltage-clamped boundary conditions are solved consistently using a boundary-element method. Based on electrostatic configurational energy, statistical-mechanical expectations of the experimentally observable relation between displaced charge and membrane voltage are predicted. Consequences of the model are investigated for variations of S4 configuration (α- and 3(10)-helical), countercharge alignment with S4 charges, protein polarizability, geometry of the gating canal, screening of S4 charges by the baths, and fixed charges located at the bath interfaces. The sliding-helix VS domain has an inherent electrostatic stability in the explored parameter space: countercharges present in the region of weak dielectric always retain an equivalent S4 charge in that region but allow sliding movements displacing 3-4 e (0). That movement is sensitive to small energy variations (<2 kT) along the path dependent on a number of electrostatic parameters tested in our simulations. These simulations show how the slope of the relation between displaced charge and voltage could be tuned in a channel.

ConGen-A Simulator-Agnostic Visual Language for Definition and Generation of Connectivity in Large and Multiscale Neural Networks.

An open challenge on the road to unraveling the brain's multilevel organization is establishing techniques to research connectivity and dynamics at different scales in time and space, as well as the links between them. This work focuses on the design of a framework that facilitates the generation of multiscale connectivity in large neural networks using a symbolic visual language capable of representing the model at different structural levels-ConGen. This symbolic language allows researchers to create and visually analyze the generated networks independently of the simulator to be used, since the visual model is translated into a simulator-independent language. The simplicity of the front end visual representation, together with the simulator independence provided by the back end translation, combine into a framework to enhance collaboration among scientists with expertise at different scales of abstraction and from different fields. On the basis of two use cases, we introduce the features and possibilities of our proposed visual language and associated workflow. We demonstrate that ConGen enables the creation, editing, and visualization of multiscale biological neural networks and provides a whole workflow to produce simulation scripts from the visual representation of the model.

Editorial: Linking experimental and computational connectomics.

Large-scale in silico experimentation depends on the generation of connectomes beyond available anatomical structure. We suggest that linking research across the fields of experimental connectomics, theoretical neuroscience, and high-performance computing can enable a new generation of models bridging the gap between biophysical detail and global function. This Focus Feature on "Linking Experimental and Computational Connectomics" aims to bring together some examples from these domains as a step toward the development of more comprehensive generative models of multiscale connectomes.

Toward Rigorous Parameterization of Underconstrained Neural Network Models Through Interactive Visualization and Steering of Connectivity Generation.

Simulation models in many scientific fields can have non-unique solutions or unique solutions which can be difficult to find. Moreover, in evolving systems, unique final state solutions can be reached by multiple different trajectories. Neuroscience is no exception. Often, neural network models are subject to parameter fitting to obtain desirable output comparable to experimental data. Parameter fitting without sufficient constraints and a systematic exploration of the possible solution space can lead to conclusions valid only around local minima or around non-minima. To address this issue, we have developed an interactive tool for visualizing and steering parameters in neural network simulation models. In this work, we focus particularly on connectivity generation, since finding suitable connectivity configurations for neural network models constitutes a complex parameter search scenario. The development of the tool has been guided by several use cases-the tool allows researchers to steer the parameters of the connectivity generation during the simulation, thus quickly growing networks composed of multiple populations with a targeted mean activity. The flexibility of the software allows scientists to explore other connectivity and neuron variables apart from the ones presented as use cases. With this tool, we enable an interactive exploration of parameter spaces and a better understanding of neural network models and grapple with the crucial problem of non-unique network solutions and trajectories. In addition, we observe a reduction in turn around times for the assessment of these models, due to interactive visualization while the simulation is computed.

Code Generation in Computational Neuroscience: A Review of Tools and Techniques.

Advances in experimental techniques and computational power allowing researchers to gather anatomical and electrophysiological data at unprecedented levels of detail have fostered the development of increasingly complex models in computational neuroscience. Large-scale, biophysically detailed cell models pose a particular set of computational challenges, and this has led to the development of a number of domain-specific simulators. At the other level of detail, the ever growing variety of point neuron models increases the implementation barrier even for those based on the relatively simple integrate-and-fire neuron model. Independently of the model complexity, all modeling methods crucially depend on an efficient and accurate transformation of mathematical model descriptions into efficiently executable code. Neuroscientists usually publish model descriptions in terms of the mathematical equations underlying them. However, actually simulating them requires they be translated into code. This can cause problems because errors may be introduced if this process is carried out by hand, and code written by neuroscientists may not be very computationally efficient. Furthermore, the translated code might be generated for different hardware platforms, operating system variants or even written in different languages and thus cannot easily be combined or even compared. Two main approaches to addressing this issues have been followed. The first is to limit users to a fixed set of optimized models, which limits flexibility. The second is to allow model definitions in a high level interpreted language, although this may limit performance. Recently, a third approach has become increasingly popular: using code generation to automatically translate high level descriptions into efficient low level code to combine the best of previous approaches. This approach also greatly enriches efforts to standardize simulator-independent model description languages. In the past few years, a number of code generation pipelines have been developed in the computational neuroscience community, which differ considerably in aim, scope and functionality. This article provides an overview of existing pipelines currently used within the community and contrasts their capabilities and the technologies and concepts behind them.

Cation Selectivity in Biological Cation Channels Using Experimental Structural Information and Statistical Mechanical Simulation.

Cation selective channels constitute the gate for ion currents through the cell membrane. Here we present an improved statistical mechanical model based on atomistic structural information, cation hydration state and without tuned parameters that reproduces the selectivity of biological Na+ and Ca2+ ion channels. The importance of the inclusion of step-wise cation hydration in these results confirms the essential role partial dehydration plays in the bacterial Na+ channels. The model, proven reliable against experimental data, could be straightforwardly used for designing Na+ and Ca2+ selective nanopores.

Integrating Visualizations into Modeling NEST Simulations.

Modeling large-scale spiking neural networks showing realistic biological behavior in their dynamics is a complex and tedious task. Since these networks consist of millions of interconnected neurons, their simulation produces an immense amount of data. In recent years it has become possible to simulate even larger networks. However, solutions to assist researchers in understanding the simulation's complex emergent behavior by means of visualization are still lacking. While developing tools to partially fill this gap, we encountered the challenge to integrate these tools easily into the neuroscientists' daily workflow. To understand what makes this so challenging, we looked into the workflows of our collaborators and analyzed how they use the visualizations to solve their daily problems. We identified two major issues: first, the analysis process can rapidly change focus which requires to switch the visualization tool that assists in the current problem domain. Second, because of the heterogeneous data that results from simulations, researchers want to relate data to investigate these effectively. Since a monolithic application model, processing and visualizing all data modalities and reflecting all combinations of possible workflows in a holistic way, is most likely impossible to develop and to maintain, a software architecture that offers specialized visualization tools that run simultaneously and can be linked together to reflect the current workflow, is a more feasible approach. To this end, we have developed a software architecture that allows neuroscientists to integrate visualization tools more closely into the modeling tasks. In addition, it forms the basis for semantic linking of different visualizations to reflect the current workflow. In this paper, we present this architecture and substantiate the usefulness of our approach by common use cases we encountered in our collaborative work.

Voltage sensing in ion channels: mesoscale simulations of biological devices.

Electrical signaling via voltage-gated ion channels depends upon the function of a voltage sensor (VS), identified with the S1-S4 domain in voltage-gated K(+) channels. Here we investigate some energetic aspects of the sliding-helix model of the VS using simulations based on VS charges, linear dielectrics, and whole-body motion. Model electrostatics in voltage-clamped boundary conditions are solved using a boundary element method. The statistical mechanical consequences of the electrostatic configurational energy are computed to gain insight into the sliding-helix mechanism and to predict experimentally measured ensemble properties such as gating charge displaced by an applied voltage. Those consequences and ensemble properties are investigated for two alternate S4 configurations, α and 3(10) helical. Both forms of VS are found to have an inherent electrostatic stability. Maximal charge displacement is limited by geometry, specifically the range of movement where S4 charges and countercharges overlap in the region of weak dielectric. Charge displacement responds more steeply to voltage in the α-helical than in the 3(10)-helical sensor. This difference is due to differences on the order of 0.1 eV in the landscapes of electrostatic energy. As a step toward integrating these VS models into a full-channel model, we include a hypothetical external load in the Hamiltonian of the system and analyze the energetic input-output relation of the VS.

Voltage sensor of ion channels and enzymes.

Placed in the cell membrane (a two-dimensional environment), ion channels and enzymes are able to sense voltage. How these proteins are able to detect the difference in the voltage across membranes has attracted much attention, and at times, heated debate during the last few years. Sodium, Ca2+ and K+ voltage-dependent channels have a conserved positively charged transmembrane (S4) segment that moves in response to changes in membrane voltage. In voltage-dependent channels, S4 forms part of a domain that crystallizes as a well-defined structure consisting of the first four transmembrane (S1-S4) segments of the channel-forming protein, which is defined as the voltage sensor domain (VSD). The VSD is tied to a pore domain and VSD movements are allosterically coupled to the pore opening to various degrees, depending on the type of channel. How many charges are moved during channel activation, how much they move, and which are the molecular determinants that mediate the electromechanical coupling between the VSD and the pore domains are some of the questions that we discuss here. The VSD can function, however, as a bona fide proton channel itself, and, furthermore, the VSD can also be a functional part of a voltage-dependent phosphatase.

Relating microscopic charge movement to macroscopic currents: the Ramo-Shockley theorem applied to ion channels.

Since the discovery of gating current, electrophysiologists have studied the movement of charged groups within channel proteins by changing potential and measuring the resulting capacitive current. The relation of atomic-scale movements of charged groups to the gating current measured in an external circuit, however, is not obvious. We report here that a general solution to this problem exists in the form of the Ramo-Shockley theorem. For systems with different amounts of atomic detail, we use the theorem to calculate the gating charge produced by movements of protein charges. Even without calculation or simulation, the Ramo-Shockley theorem eliminates a class of interpretations of experimental results. The theorem may also be used at each time step of simulations to compute external current.