Pavol Bauer

Efficient Inter-Process Synchronization for Parallel Discrete Event Simulation on Multicores

We present a new technique for controlling optimism in Parallel Discrete Event Simulation on multicores. It is designed to be suitable for simulating models, in which the time intervals between successive events between different processes are highly variable, and have no lower bounds. In our technique, called Dynamic Local Time Window Estimates (DLTWE), each processor communicates time estimates of its next inter-processor event to (some of) its neighbors, which use the estimates as bounds for advancement of their local simulation time. We have implemented our technique in a parallel simulator for simulation of spatially extended Markovian processes of interacting entities, which can model chemical reactions, processes from biology, epidemics, and many other applications. Intervals between successive events are exponentially distributed, thus having a significant variance and no lower bound. We show that the DLTWE technique can be tuned to drastically reduce the frequency of rollbacks and enable speedups which is superior to that obtained by other works. We also show that the DLTWE technique significantly improves performance over other existing techniques for optimism control that attempt to predict arrival of inter-process events by statistical techniques.

Sensitivity estimation and inverse problems in spatial stochastic models of chemical kinetics

Sensitivity estimation and inverse problems in spatial stochastic models of chemical kinetics

Data-driven network modelling of disease transmission using complete population movement data: spread of VTEC O157 in Swedish cattle.

European Union legislation requires member states to keep national databases of all bovine animals. This allows for disease spread models that includes the time-varying contact network and population demographic. However, performing data-driven simulations with a high degree of detail are computationally challenging. We have developed an efficient and flexible discrete-event simulator SimInf for stochastic disease spread modelling that divides work among multiple processors to accelerate the computations. The model integrates disease dynamics as continuous-time Markov chains and livestock data as events. In this study, all Swedish livestock data (births, movements and slaughter) from July 1st 2005 to December 31st 2013 were included in the simulations. Verotoxigenic Escherichia coli O157:H7 (VTEC O157) are capable of causing serious illness in humans. Cattle are considered to be the main reservoir of the bacteria. A better understanding of the epidemiology in the cattle population is necessary to be able to design and deploy targeted measures to reduce the VTEC O157 prevalence and, subsequently, human exposure. To explore the spread of VTEC O157 in the entire Swedish cattle population during the period under study, a within- and between-herd disease spread model was used. Real livestock data was incorporated to model demographics of the population. Cattle were moved between herds according to real movement data. The results showed that the spatial pattern in prevalence may be due to regional differences in livestock movements. However, the movements, births and slaughter of cattle could not explain the temporal pattern of VTEC O157 prevalence in cattle, despite their inherently distinct seasonality.

A role for solute carrier family 10 member 4, or vesicular aminergic‐associated transporter, in structural remodelling and transmitter release at the mouse neuromuscular junction

The solute carrier and presynaptic vesicle protein solute carrier family 10 member 4, or vesicular aminergic‐associated transporter (VAAT), was recently proven to have a modulatory role in central cholinergic signalling. It is currently unknown whether VAAT also affects peripheral cholinergic synapses. Here we demonstrated a regulatory role for the presynaptic vesicle protein VAAT in neuromuscular junction (NMJ) development and function. NMJs lacking VAAT had fewer branch points, whereas endplates showed an increased number of islands. Whereas the amplitude of spontaneous miniature endplate potentials in VAAT‐deficient NMJs was decreased, the amplitude of evoked endplate potentials and the size of the readily releasable pool of vesicles were both increased. Moreover, VAAT‐deficient NMJs displayed aberrant short‐term synaptic plasticity with enhanced synaptic depression in response to high‐frequency stimulation. Finally, the transcript levels of cholinergic receptor subunits in VAAT‐deficient muscles were increased, indicating a compensatory postsynaptic sensitization. Our results suggested that VAAT modulates NMJ transmission efficiency and, as such, may represent a novel target for treatment of disorders affecting motor neurons.

Fine-Grained Local Dynamic Load Balancing in PDES

We present a fine-grained load migration protocol intended for parallel discrete event simulation (PDES) of spatially extended models. Typical models have domains that are fine-grained discretizations of some volume, e.g., a cell, using an irregular three-dimensional mesh, where most events span several voxels. Phenomena of interest in, e.g., cellular biology, are often non-homogeneous and migrate over the simulated domain, making load balancing a crucial part of a successful PDES. Our load migration protocol is local in the sense that it involves only those processors that exchange workload, and does not affect the running parallel simulation. We present a detailed description of the protocol and a thorough proof for its correctness. We combine our protocol with a strategy for deciding when and what load to migrate, which optimizes both for load balancing and inter-processor communication using tunable parameters. Our evaluation shows that the overhead of the load migration protocol is negligible, and that it significantly reduces the number of rollbacks caused by load imbalance. On the other hand, the implementation mechanisms that we added to support fine-grained load balancing incur a significant cost.

Ventral hippocampal OLM cells control type 2 theta oscillations and response to predator odor

Dorsal and ventral hippocampus regions exert cognition and emotion-related functions, respectively. Since both regions display rhythmic activity, specific neural oscillatory pacemakers may underlie their functional dichotomy. Type 1 theta oscillations are independent of cholinergic transmission and are observed in the dorsal hippocampus during movement and exploration. In contrast, type 2 theta depends on acetylcholine and appears when animals are exposed to emotionally laden contexts such as a predator presence. Despite its involvement in emotions, type 2 theta has not been associated with the ventral hippocampus. Here, we show that optogenetic activation of oriens-lacunosum moleculare (OLM) interneurons in the ventral hippocampus drives type 2 theta. Moreover, we found that type 2 theta generation is associated with increased risk-taking behavior in response to predator odor. These results demonstrate that two theta oscillations subtypes originate in the two hippocampal regions that predominantly underlie either cognitive or emotion-related functions.There are two subtypes of hippocampal theta oscillations that differ in frequency range, pharmacology, and behavioural correlates. Here, the authors report that activity of OLM interneurons in the ventral hippocampus mediates type 2 theta, associated with increased risk-taking in the presence of predator threat.

Multiscale modelling via split-step methods in neural firing

ABSTRACT Neuronal models based on the Hodgkin–Huxley equation form a fundamental framework in the field of computational neuroscience. While the neuronal state is often modelled deterministically, experimental recordings show stochastic fluctuations, presumably driven by molecular noise from the underlying microphysical conditions. In turn, the firing of individual neurons gives rise to an electric field in extracellular space, also thought to affect the firing pattern of nearby neurons. We develop a multiscale model which combines a stochastic ion channel gating process taking place on the neuronal membrane, together with the propagation of an action potential along the neuronal structure. We also devise a numerical method relying on a split-step strategy which effectively couples these two processes and we experimentally test the feasibility of this approach. We finally also explain how the approach can be extended with Maxwell’s equations to allow the potential to be propagated in extracellular space.