Underlying Network Connectivity Research Articles

Photochemical Control of Network Topology in PEG Hydrogels.

Hydrogels are often synthesized through photoinitiated step-, chain-, and mixed-mode polymerizations, generating diverse network topologies and resultant material properties that depend on the underlying network connectivity. While many photocrosslinking reactions are available, few afford controllable connectivity of the hydrogel network. Herein, a versatile photochemical strategy is introduced for tuning the structure of poly(ethylene glycol) (PEG) hydrogels using macromolecular monomers functionalized with maleimide and styrene moieties. Hydrogels are prepared along a gradient of topologies by varying the ratio of step-growth (maleimide dimerization) to chain-growth (maleimide-styrene alternating copolymerization) network-forming reactions. The initial PEG content and final network physical properties (e.g., modulus, swelling, diffusivity) are tailored in an independent manner, highlighting configurable gel mechanics and reactivity. These photochemical reactions allow high-fidelity photopatterning and 3D printing and are compatible with 2D and 3D cell culture. Ultimately, this photopolymer chemistry allows facile control over network connectivity to achieve adjustable material properties for broad applications.

The Neurostimulationist will see you now: prescribing direct electrical stimulation therapies for the human brain in epilepsy and beyond.

As the pace of research in implantable neurotechnology increases, it is important to take a step back and see if the promise lives up to our intentions. While direct electrical stimulation applied intracranially has been used for the treatment of various neurological disorders, such as Parkinson's, epilepsy, clinical depression, and Obsessive-compulsive disorder, the effectiveness can be highly variable. One perspective is that the inability to consistently treat these neurological disorders in a standardized way is due to multiple, interlaced factors, including stimulation parameters, location, and differences in underlying network connectivity, leading to a trial-and-error stimulation approach in the clinic. An alternate view, based on a growing knowledge from neural data, is that variability in this input (stimulation) and output (brain response) relationship may be more predictable and amenable to standardization, personalization, and, ultimately, therapeutic implementation. In this review, we assert that the future of human brain neurostimulation, via direct electrical stimulation, rests on deploying standardized, constrained models for easier clinical implementation and informed by intracranial data sets, such that diverse, individualized therapeutic parameters can efficiently produce similar, robust, positive outcomes for many patients closer to a prescriptive model. We address the pathway needed to arrive at this future by addressing three questions, namely: (1) why aren't we already at this prescriptive future?; (2) how do we get there?; (3) how far are we from this Neurostimulationist prescriptive future? We first posit that there are limited and predictable ways, constrained by underlying networks, for direct electrical stimulation to induce changes in the brain based on past literature. We then address how identifying underlying individual structural and functional brain connectivity which shape these standard responses enable targeted and personalized neuromodulation, bolstered through large-scale efforts, including machine learning techniques, to map and reverse engineer these input-output relationships to produce a good outcome and better identify underlying mechanisms. This understanding will not only be a major advance in enabling intelligent and informed design of neuromodulatory therapeutic tools for a wide variety of neurological diseases, but a shift in how we can predictably, and therapeutically, prescribe stimulation treatments the human brain.

Network inference combining mutual information rate and statistical tests

In this paper, we present a method that combines information-theoretical and statistical approaches to infer connectivity in complex networks using time-series data. The method is based on estimations of the Mutual Information Rate for pairs of time-series and on statistical significance tests for connectivity acceptance using the false discovery rate method for multiple hypothesis testing. We provide the mathematical background on Mutual Information Rate, discuss the statistical significance tests and the false discovery rate. Further on, we present results for correlated normal-variates data, coupled circle and coupled logistic maps, coupled Lorenz systems and coupled stochastic Kuramoto phase oscillators. Following up, we study the effect of noise on the presented methodology in networks of coupled stochastic Kuramoto phase oscillators and of coupling heterogeneity degree on networks of coupled circle maps. We show that the method can infer the correct number and pairs of connected nodes, by means of receiver operating characteristic curves. In the more realistic case of stochastic data, we demonstrate its ability to infer the structure of the initial connectivity matrices. The method is also shown to recover the initial connectivity matrices for dynamics on the nodes of Erdős-Rényi and small-world networks with varying coupling heterogeneity in their connections. The highlight of the proposed methodology is its ability to infer the underlying network connectivity based solely on the recorded datasets.

Changes in functional connectivity associated with facial expression processing over the working adult lifespan

The recognition of negative emotions from facial expressions is shown to decline across the adult lifespan, with some evidence that this decline begins around middle age. While some studies have suggested ageing may be associated with changes in neural response to emotional expressions, it is not known whether ageing is associated with changes in the network connectivity associated with processing emotional expressions. In this study, we examined the effect of participant age on whole-brain connectivity to various brain regions that show connectivity during emotion processing, namely, the left and right amygdalae, medial prefrontal cortex (mPFC), and right posterior superior temporal sulcus (rpSTS). The study involved healthy participants aged 20–65 years who completed an implicit affect processing task involving facial expressions displaying anger, fear, happiness, and neutral expressions during functional magnetic resonance imaging (fMRI). Participants were also tested on recognition accuracy during an emotion labelling task. We found that participant age was negatively associated with connectivity between the left amygdala and voxels in the left occipital pole; between the rpSTS and voxels in the orbitofrontal cortex; and between the mPFC and cingulate cortex and left insular cortex. Furthermore, these effects were due to a greater age-related decline in brain connectivity for negative expressions compared to happy and neutral expressions. There was, however, no significant relationship between age and emotion recognition accuracy (though in the expected direction), and between connectivity strength and emotion recognition accuracy. Together, these results provide evidence for a specific age-related decline in the neural processing of negative emotions, and could suggest that changes in underlying network connectivity over the working adult lifespan might occur before any marked decline in the recognition of the emotion.

Reconstruction of network connectivity by the interplay between complex structure and dynamics to discover climate networks

Characterizing network connectivity from measures of correlation and causality is of utmost importance to study the Earth’s climate system through the statistical analysis of collective dynamics. A major challenge is to infer functional connectivity from the observations, and different methods were applied to linear and nonlinear dynamics simulated from different complex topologies widely observed in empirical systems to quantify the limits on reconstruction of functional connectivity. Climate networks were constructed from the global air temperature on the grid points of surface and pressure level of 850 hPa in monthly time step based on the robust statistical analysis with the null hypothesis of uncorrelated dynamics. Mutual information method was chosen after examining on the artificial cases consisting of desirable features of the climate system utilizing complex networks. Analyzing imbalanced results persuaded further investigation as cross-validation to have a satisfied comparison against ground truth with the employed statistical tools. The results indicated that the interplay between the structure and dynamics led to very important differences in reconstructing the underlying network connectivity, especially when different methods were considered, even under the most rigorous statistical tests. Analyzing structural characteristics of the constructed climate network on the surface level unraveled large-scale climate oscillations with the significant connectivities over the oceans. It was also noticed that the constructed climate network on the pressure level revealed more extended connectivities providing useful information to identify underlying physical interactions between ocean and atmosphere with coupling pattern on different pressure levels.

Characterizing controllable subspace and herdability of signed weighted networks via graph partition

Herdability is a variant of controllability, and is an indicator of the ability to drive system states to a specific subset of the state space. This paper characterizes the controllable subspace and herdability of signed weighted networks. Specifically, a dynamic signed leader–follower network is considered, in which a small subset of the network nodes (i.e., the leaders) is endowed with exogenous control input and the remaining nodes are influenced by the leaders via the underlying network connectivity. The considered network permits positive and negative edges to capture cooperative and competitive interactions, resulting in a signed graph. Motivated by practical application, the system states are required to be driven by the leaders to be element-wise above a positive threshold, i.e., a specific subset rather than the entire state space as in classical controllability. Graph partitions are exploited to characterize the controllable subspace of the system, from which sufficient conditions are derived to render the system herdable. It is revealed that the quotient graph can be used to infer the herdability of the original graph, wherein criteria of the herdability of quotient graphs are developed based on positive systems. Examples are provided to illustrate the developed topological characterizations.

Functional human brain connectivity during labor and its alteration under epidural analgesia.

This study used functional magnetic resonance imaging to explore the neural networks of pain during labor and its relief. It was hypothesized that epidural analgesia would affect the neural activities and the underlying network connectivity. Analysis using dynamic causal modelling and functional connectivity was performed to investigate the spatial activity and network connection of labor pain with and without epidural analgesia. This Institutional Review Board approved study acquired Magnetic Resonance Imaging from 15 healthy women of spontaneous normal delivery (with/without epidural analgesia = 7/8, aged 29.6 ± 2.3 and 29.3 ± 4.8years old respectively) using a 1.5 Tesla scanner. Numerical rating score of pain was evaluated by a research nurse in the beginning of the first stage of labor and approximately 30min after imaging examination. Six regions of interested from the activated clusters and literature were selected for dynamic causal modelling, which included primary and secondary somatosensory cortex, middle frontal gyrus, anterior cingulate cortex, insula and lentiform. Functional connectivity was calculated from selected sensory and affective regions. All analyses were performed by using software of statistical parametric mapping version 8 and CONN functional connectivity toolbox. The result showed that the experience of labor pain can lead to activations within a distributed brain network. The pain relief from epidural analgesia can be accompanied with altered functional connectivity, which was most evident in the cingulo-frontal system. The present study, therefore, provides an overview of a pain-related neural network that occur during labor and upon epidural analgesia.

In vivo measurement of brain network connectivity reflects progression and intrinsic disease severity in a model of temporal lobe epilepsy

Different types of brain injury, such as status epilepticus (SE), trauma, or stroke may initiate the process of epileptogenesis and lead to the development of temporal lobe epilepsy. Epileptogenesis is characterized by an initial latent period during which impaired network communication and synaptic circuit alterations are occurring. Ultimately, these modifications result in the development of spontaneous recurrent seizures (SRS). Current knowledge on the functional connectivity network changes during epileptogenesis and how network alterations relate to seizure is very limited. To investigate these underlying network connectivity modifications, we imaged epileptic and control rats by means of resting-state functional MRI (rsfMRI) during epileptogenesis. A cohort of animals was video-electroencephalography (video-EEG) monitored continuously over 12 weeks to determine disease severity during the course of disease, with the first SRS appearing around 2 weeks post-SE for most of the animals. Epileptic animals displayed a significant wide-spread hyposynchrony at 2 weeks post-SE, followed by a significant increase in network synchronicity from 2 to 4 weeks post-SE. Interestingly, subjects with a delayed epilepsy onset demonstrated significantly lower synchronicity compared to controls and the epileptic group at 4 weeks post-SE. Finally, network connectivity at 4 weeks post-SE was found to correlate with seizure onset (r = 0.858, p < .0001) and disease severity measured over 12 weeks (e.g. cingulate cortex: r = 0.863, p = .002), suggesting a possible network strengthening upon seizure reoccurrence. Our findings indicate that epileptogenesis is characterized by an initial hyposynchrony of brain networks and the disease-associated progression reflects disease severity.

A Complex Network Framework to Model Cognition: Unveiling Correlation Structures from Connectivity

Several approaches to cognition and intelligence research rely on statistics-based model testing, namely, factor analysis. In the present work, we exploit the emerging dynamical system perspective putting the focus on the role of the network topology underlying the relationships between cognitive processes. We go through a couple of models of distinct cognitive phenomena and yet find the conditions for them to be mathematically equivalent. We find a nontrivial attractor of the system that corresponds to the exact definition of a well-known network centrality and hence stresses the interplay between the dynamics and the underlying network connectivity, showing that both of the two are relevant. Correlation matrices evince there must be a meaningful structure underlying real data. Nevertheless, the true architecture regarding the connectivity between cognitive processes is still a burning issue of research. Regardless of the network considered, it is always possible to recover a positive manifold of correlations. Furthermore, we show that different network topologies lead to different plausible statistical models concerning the correlation structure, ranging from one to multiple factor models and richer correlation structures.

Hyperconnected molecular glass network architectures with exceptional elastic properties

Hyperconnected network architectures can endow nanomaterials with remarkable mechanical properties that are fundamentally controlled by designing connectivity into the intrinsic molecular structure. For hybrid organic–inorganic nanomaterials, here we show that by using 1,3,5 silyl benzene precursors, the connectivity of a silicon atom within the network extends beyond its chemical coordination number, resulting in a hyperconnected network with exceptional elastic stiffness, higher than that of fully dense silica. The exceptional intrinsic stiffness of these hyperconnected glass networks is demonstrated with molecular dynamics models and these model predictions are calibrated through the synthesis and characterization of an intrinsically porous hybrid glass processed from 1,3,5(triethoxysilyl)benzene. The proposed molecular design strategy applies to any materials system wherein the mechanical properties are controlled by the underlying network connectivity.

Epidemic potential by sexual activity distributions.

For sexually transmitted infections like HIV to propagate through a population, there must be a path linking susceptible cases to currently infectious cases. The existence of such paths depends in part on the degree distribution. Here, we use simulation methods to examine how two features of the degree distribution affect network connectivity: Mean degree captures a volume dimension, while the skewness of the upper tail captures a shape dimension. We find a clear interaction between shape and volume: When mean degree is low, connectivity is greater for long-tailed distributions, but at higher mean degree, connectivity is greater in short-tailed distributions. The phase transition to a giant component and giant bicomponent emerges as a positive function of volume, but it rises more sharply and ultimately reaches more people in short-tail distributions than in long-tail distributions. These findings suggest that any interventions should be attuned to how practices affect both the volume and shape of the degree distribution, noting potential unanticipated effects. For example, policies that primarily affect high-volume nodes may not be effective if they simply redistribute volume among lower degree actors, which appears to exacerbate underlying network connectivity.

Estimation of Directed Effective Connectivity from fMRI Functional Connectivity Hints at Asymmetries of Cortical Connectome.

The brain exhibits complex spatio-temporal patterns of activity. This phenomenon is governed by an interplay between the internal neural dynamics of cortical areas and their connectivity. Uncovering this complex relationship has raised much interest, both for theory and the interpretation of experimental data (e.g., fMRI recordings) using dynamical models. Here we focus on the so-called inverse problem: the inference of network parameters in a cortical model to reproduce empirically observed activity. Although it has received a lot of interest, recovering directed connectivity for large networks has been rather unsuccessful so far. The present study specifically addresses this point for a noise-diffusion network model. We develop a Lyapunov optimization that iteratively tunes the network connectivity in order to reproduce second-order moments of the node activity, or functional connectivity. We show theoretically and numerically that the use of covariances with both zero and non-zero time shifts is the key to infer directed connectivity. The first main theoretical finding is that an accurate estimation of the underlying network connectivity requires that the time shift for covariances is matched with the time constant of the dynamical system. In addition to the network connectivity, we also adjust the intrinsic noise received by each network node. The framework is applied to experimental fMRI data recorded for subjects at rest. Diffusion-weighted MRI data provide an estimate of anatomical connections, which is incorporated to constrain the cortical model. The empirical covariance structure is reproduced faithfully, especially its temporal component (i.e., time-shifted covariances) in addition to the spatial component that is usually the focus of studies. We find that the cortical interactions, referred to as effective connectivity, in the tuned model are not reciprocal. In particular, hubs are either receptors or feeders: they do not exhibit both strong incoming and outgoing connections. Our results sets a quantitative ground to explore the propagation of activity in the cortex.

Investigating intrinsic and evoked activities in cultured neuronal networks by dimensional reduction techniques and high-density MEAs

High density microelectrode arrays (MEAs) provide extracellular recordings from thousands of closely spaced electrodes and with sub-millisecond resolution. These devices offer thus novel capabilities to investigate the interplay between ongoing and evoked electrophysiological signaling within networks in-vitro. However, to effectively take advantage from the spatiotemporal resolution of these MEAs, adapted analysis tools are needed. Here we report on our recent advancements toward this goal. A novel high density MEA with on-chip stimulating electrodes was used to record from 4096 electrodes and electrical stimulation was delivered alternatively through one of the 16 equally spaced electrodes on hippocampal primary cultures derived from mouse. The evoked activities propagated reliably across the network and were specific to the stimulating electrode [1]. Moreover, from the second week in vitro cell cultures also displayed synchronous like events (called synchronous burst events, SBEs) that propagated similarly to the evoked activities across the entire network. These propagations might be informative of the underlying network connectivity and their classification based on the spatiotemporal patterns might elucidate the network's organization and its ongoing dynamic. Former attempts, in classifying SBEs, considered simplified descriptors of neural activity (e.g. center activity trajectory [2]). Here, we have adopted a more rigorous approach by applying dimensional reduction techniques (PCA) that take advantage of the redundancy and of the sparseness of multi-unit recordings. We found that a large fraction (i.e. >50%) of the variance of the network events (either evoked or spontaneous) was explained by as few as 3 principal components (PCs). By increasing the number of PCs (i.e. ~10) up to 80% of the total variance of the network events could be explained. Thus, the PCA approach constitutes an effective methodology to represent the spontaneous/evoked events in a lower dimensional space. As a consequence the principled PCA methodology we developed improved the clustering of the network events respect to existing methodologies [2] in different experimental settings (e.g. treatment with chemical compounds). Finally, the PCA clustering approach also allowed to infer on state dependent processing phenomena occurring in these networks.

Uncovering the structure and temporal dynamics of information propagation

AbstractTime plays an essential role in the diffusion of information, influence, and disease over networks. In many cases we can only observe when a node is activated by a contagion—when a node learns about a piece of information, makes a decision, adopts a new behavior, or becomes infected with a disease. However, the underlying network connectivity and transmission rates between nodes are unknown. Inferring the underlying diffusion dynamics is important because it leads to new insights and enables forecasting, as well as influencing or containing information propagation. In this paper we model diffusion as a continuous temporal process occurring at different rates over a latent, unobserved network that may change over time. Given information diffusion data, we infer the edges and dynamics of the underlying network. Our model naturally imposes sparse solutions and requires no parameter tuning. We develop an efficient inference algorithm that uses stochastic convex optimization to compute online estimates of the edges and transmission rates. We evaluate our method by tracking information diffusion among 3.3 million mainstream media sites and blogs, and experiment with more than 179 million different instances of information spreading over the network in a one-year period. We apply our network inference algorithm to the top 5,000 media sites and blogs and report several interesting observations. First, information pathways for general recurrent topics are more stable across time than for on-going news events. Second, clusters of news media sites and blogs often emerge and vanish in a matter of days for on-going news events. Finally, major events, for example, large scale civil unrest as in the Libyan civil war or Syrian uprising, increase the number of information pathways among blogs, and also increase the network centrality of blogs and social media sites.

Investigating the interplay between intrinsic and evoked activities in cultured neuronal networks by dimensional reduction techniques and high-density MEAs

High density microelectrode arrays (MEAs) provide extracellular recordings from thousand of electrodes (http://www.3brain.com) and offer novel capabilities to investigate electrophysiological signaling in cultured neuronal networks and in ex vivo brain tissues. In this study we report on our recent technological and data analysis advancements to investigate the propagation and the interplay of spontaneous and electrically evoked activities in cultured networks. To do so, a novel high density MEA with on-chip stimulating electrodes was realized. It provides whole-array recordings from 4096 electrodes (pitch of 81 um, active area of 8 mm by 8 mm) and electrical stimulation from 16 electrodes located every 8 recording sites. Here, this device was used to interface hippocampal neuronal networks. From the second week in vitro [1] these cultures display a peculiar intrinsic firing regime characterized by periodic synchronized network-wide bursts. These network bursts originate from specific ignition sites of the cultured network (i.e. characterized by more excitable cells) and can propagate through the entire network. These propagations are informative of the underlying network connectivity and their classification based on their spatiotemporal patterns might elucidate the network's organization and its ongoing dynamic. Previous studies [1] have described the trajectories of these propagations by tracking their center of activity trajectory (CAT). Although CATs provide a good overall description of spatiotemporal patterns, they are not suited for fine studies on these propagations. Here we have adopted a more rigorous approach by applying dimensional reduction techniques that take advantage of the redundancy and of the sparseness of multi-unit recordings. Our results show that by Principal Component Analysis (PCA) we can properly reconstruct the time course of spatiotemporal propagations with a minimal set of three components (i.e. explaining ~90% of the variance of the trajectories). The PCA classification based approach allowed us to characterize both intrinsic and electrically evoked network-wide propagations. Interestingly, our results show that electrical stimulation can evoke (i) distinctive propagations that depend on the specific spatial-temporal properties of the stimulus as well as (ii) propagations that are already expressed in the intrinsic 'repertoire' of network bursts. Here we will discuss these results and our observations on the interplay between intrinsic and electrically evoked network-wide propagations.

Pairwise correlation graphs from hippocampal population activity have highly non-random, low-dimensional clique topology

We analyzed the structure of cliques in pairwise correlation graphs obtained from population activity in hippocampus under a variety of behavioral conditions: open field exploration, wheel running, and sleep. Using topological data analysis, we found these graphs to be highly non-random by comparing their clique structure to that of Erdos-Renyi random graphs with matching edge probabilities. The clique topology in all three conditions was low-dimensional and consistent with what is theoretically expected from place cells during spatial navigation. Remarkably, the same pattern was also observed in the sleep data. To better understand how this correlation structure might arise, we considered a recurrent network with a simple Hebbian learning rule. We found the same clique structure emerged in the network when recurrent connections were formed in the presence of random sparse patterns of activity. This may provide a generic mechanism explaining how low-dimensional clique topology arises in hippocampal data. Pairwise correlations are an important tool for understanding neuronal population activity [1]. Pairwise correlation graphs, where the edges reflect high levels of correlation between neurons, are often used as a proxy for underlying network connectivity. In this work, we are motivated by the question: What is the structure of pairwise correlations in hippocampal population activity? To address this question, we analyzed graded families of pairwise correlation graphs, parametrized by the threshold on correlation strength used to define the edges. In brain areas with receptive fields or place fields, the structure of the neural code has strong implications for structure of cliques in these graphs. This can be detected by examining the clique topology of the graph (specifically, topological invariants called 'Betti numbers' of the clique complex), and is closely tied to the dimension and topology of the underlying space [2]. For example, for hippocampal place cell activity during spatial exploration, pairwise correlation graphs are expected to have highly non-random and low-dimensional clique topology, due to the arrangement of place fields in a low-dimensional environment. To test this hypothesis, we analyzed pairwise correlations graphs inferred from hippocampal CA1 data under a variety of behavioral conditions (open field exploration, wheel running, and sleep). We found that pairwise correlation graphs are (1) highly non-random, in a way that is easily detected by the clique topology; and (2) the clique topology is low-dimensional and consistent with what is expected of neurons firing according to place fields. Remarkably, the sleep data showed a similar pattern, despite the fact that place fields are not observed under these conditions. This led us to ask: How might low-dimensional clique topology arise in a highly plastic network, like hippocampus, during non-spatial behaviors? To answer this question, we investigated a simple recurrent network, where pairwise connections are strengthened by Hebbian learning in the presence of sparse random patterns of externally driven activity. We found that the resulting network had highly non-random and low-dimensional clique topology, similar to what was observed in hippocampal data. This suggests that low-dimensional clique structure in pairwise correlation graphs may naturally emerge in networks with Hebbian plasticity.

Computing Along Routes via Gossiping

In this paper, we study a class of distributed computing problems where a group of nodes (destinations) is interested in a function of data which are stored by another group of nodes (sources). We assume that the function of interest is separable, i.e., it can be decomposed as a sum of functions of local variables, a case that subsumes several interesting types of queries. One approach to solve this problem is to route raw information from the sources to the interested destinations by either unicasting or multicasting. The second approach is to compute the function of interest along some routes while propagating the information from the sources to the destinations. Considering the second scenario, the goal of this paper is to examine how information should be forwarded to the intended recipients, computing along the routes by gossiping with selected neighbors. Unlike efficient unicasting/multicasting problems, nodes are interested in a specific function of the data, rather than the raw data themselves. Moreover, unlike standard gossiping problems, in our case, the information needs to flow in a specific direction. Given the underlying network connectivity and the source-destination sets, we provide necessary and sufficient conditions on the update weights (referred to as codes) so that the destination nodes converge to the desired function of the source values. We show that the evolution of the source states does not affect the feasibility of the problem, and we provide a detailed analysis on the spectral properties of the feasible codes. We also study the problem feasibility under some specific topologies and provide guidelines to determine infeasibility. We also formulate different strategies to design codes, and compare the performance of our solution with existing alternatives.

A Secure Approach to Distributed Internet-Enabled Metrology

This paper aims to explore ways of securely operating electrical instruments via the Internet. A system is presented, which allows a person connected to a local area network (LAN) to operate multiple electrical instruments located at other LANs via a public Web server on the Internet, regardless of firewalls, proxy servers, or source network address translators that separate them. A full system test has been performed, and data from the test are later presented. The solution is based on Microsoft .NET Remoting, which is part of the Microsoft .NET Framework. A viable application is Internet-enabled calibration, and possible users are national measurement institutes and calibration laboratories. The solution is general and does not depend on the underlying network connections involved as long as hypertext transfer protocol over secure sockets layer is supported.

Metacomputing in gigabit environments: Networks, tools, and applications

This article gives an overview over recent and current metacomputing activities of the Computing Centers at the Forschungszentrum Jülich, the GMD Forschungszentrum Informationstechnik and the University of Stuttgart. It starts with a discussion of the underlying network connections which are dedicated testbeds. A library that provides an MPI-API for metacomputing applications is presented, as well as a library that supports load-balancing and latency hiding. Finally, results from several applications ranging from tightly coupled homogeneous to loosely coupled heterogeneous metacomputing are presented.