Overcoming the Scalability Challenges of Epidemic Simulations on Blue Waters

Modeling dynamical systems represents an important application class covering a wide range of disciplines including but not limited to biology, chemistry, finance, national security, and health care. Such applications typically involve large-scale, irregular graph processing, which makes them difficult to scale due to the evolutionary nature of their workload, irregular communication and load imbalance. EpiSimdemics is such an application simulating epidemic diffusion in extremely large and realistic social contact networks. It implements a graph-based system that captures dynamics among co-evolving entities. This paper presents an implementation of EpiSimdemics in Charm++ that enables future research by social, biological and computational scientists at unprecedented data and system scales. We present new methods for application-specific processing of graph data and demonstrate the effectiveness of these methods on a Cray XE6, specifically NCSA's Blue Waters system.

Critical role of water in the direct oxidation of CO and hydrocarbons in diesel exhaust after treatment catalysis

CO and C3H6 oxidation have been carried out in the absence and presence of water over a Pd/Al2O3catalyst. It is clear that water promotes CO and, as a consequence, C3H6oxidation takes place at muchlower temperatures compared with the dry feed. The significant increase in the catalyst’s activity withrespect to CO oxidation is not simply associated with changes in surface concentration as a result ofcompetitive adsorption effects. Utilising18O2as the reactant allows the pathways whereby the oxidationdue to gaseous dioxygen and where the water activates the CO and C3H6to be distinguished. In thepresence of water, the predominant pathway is via water activation with C16O2and C16O18O being themajor species formed and oxidation with dioxygen plays a secondary role. The importance of wateractivation is further supported by the significant decrease in its effect when using D2O versus H2O.

T- and L-type Ca2+ currents in freshly dispersed smooth muscle cells from the human proximal urethra

The purpose of the present study was to characterise Ca2+ currents in smooth muscle cells isolated from biopsy samples taken from the proximal urethra of patients undergoing surgery for bladder or prostate cancer. Cells were studied at 37 degreesC using the amphotericin B perforated-patch configuration of the patch-clamp technique. Currents were recorded using Cs+-rich pipette solutions to block K+ currents. Two components of current, with electrophysiological and pharmacological properties typical of T- and L-type Ca2+ currents, were present in these cells. When steady-state inactivation curves for the L current were fitted with a Boltzmann equation, this yielded a V-1/2 of -45 +/- 5 mV. In contrast, the T current inactivated with a V-1/2 of -80 +/- 3 mV. The L currents were reduced in a concentration-dependent manner by nifedipine (ED50 = 159 +/- 54 nm) and Ni2+ (ED50 = 65 +/- 16 muM) but were enhanced when external Ca2+ was substituted with Ba2+. The T current was little affected by TTX, reduction in external Na+, application of nifedipine at concentrations below 300 nm or substitution of external Ca2+ with Ba2+, but was reduced by Ni2+ with an ED50 of 6 +/- 1 mum. When cells were stepped from -100 to -30 mV in Ca2+-free conditions, small inward currents could be detected. These were enhanced 40-fold in divalent-cation-free solution and blocked in a concentration-dependent manner by Mg2+ with an ED50 of 32 +/- 16 mum. These data support the idea that human urethral myocytes possess currents with electrophysiological and pharmacological properties typical of T- and L-type Ca2+ currents.

Extraordinary solute-stress tolerance contributes to the environmental tenacity of mycobacteria

Mycobacteria are associated with a number of well-characterized diseases, yet we know little about their stress-biology in natural ecosystems. This study focuses on the isolation and characterization of strains from Yellowstone-(YNP) and Glacier-National-Parks (GNP; USA), the majority of those identified were Mycobacterium parascrofulaceum, Mycobacterium avium (YNP) or Mycobacterium gordonae (GNP). Generally, their temperature windows for growth were >60°C; selected isolates grew at super-saturated concentrations of hydrophobic stressors and at levels of osmotic stress and chaotropic activity (up to 13.4 kJkg-1) similar to, or exceeding, those for the xerophilic fungus Aspergillus wentii and solvent-tolerant bacterium Pseudomonas putida. For example, mycobacteria grew down to 0.800 water-activity indicating that they are, with the sole exception of halophiles, more xerotolerant than other bacteria (or any Archaea). Furthermore, the fatty-acid composition of Mycobacterium cells grown over a range of salt concentrations changed less than that of other bacteria, indicating a high level of resilience, regardless of the stress load. Cells of M. parascrofulaceum, M. smegmatis and M. avium resisted the acute, potentially lethal challenges from extremes of pH (<1; >13), and saturated MgCl2-solutions (5 M; 212 kJ kg-1 chaotropicity). Collectively, these findings challenge the paradigm that bacteria have solute tolerances inferior to those of eukaryotes.

Spatio-temporal feeding selection of red deer in a mountainous landscape

Understanding the dietary consumption and selection of wild populations of generalist herbivores is hampered by the complex array of factors. Here, we determine the influence of habitat, season, and animal density, sex, and age on the diet consumption and selection of 426 red deer (Cervus elaphus scoticus) culled in Fiordland National Park, New Zealand. Our site differs from studies elsewhere both in habitat (evergreen angiosperm-dominated forests) and the intensity of hunting pressures. We predicted that deer would not consume forage in proportion to its relative availability, and that dietary consumption would change among and within years in response to hunting pressures that would also limit opportunities for age and sex segregation. Using canonical correspondence analysis, we evaluated the relative importance of different drivers of variation in diet consumption assessed from gut content and related these to available forage in the environment. We found that altitude explained the largest proportion of variation in diet consumption, reflecting the ability of deer to alter their consumption and selection in relation to their foraging grounds. Grasses formed a high proportion of the diet consumption, even for deer culled several kilometres from the alpine grasslands. In the winter months, when the alpine grasslands were largely inaccessible, less grass was eaten and deer resorted to woody plants that were avoided in the summer months. Surprisingly, there were no significant dietary differences between adults and juveniles and only subtle differences between the sexes. Sex-based differences in diet consumption are commonly observed in ungulate species and we suggest that they may have been reduced in our study area owing to decreased heterogeneity in available forage as the diversity of palatable species decreased under high deer browsing pressures, or by intense hunting pressure. © 2009 The Authors. Journal compilation © 2009 Ecological Society of Australia.

A simple approach to the supernova progenitor–explosion connection

We present a new approach to understand the landscape of supernova explosion energies, ejected nickel masses, and neutron star birth masses. In contrast to other recent parametric approaches, our model predicts the properties of neutrino-driven explosions based on the pre-collapse stellar structure without the need for hydrodynamic simulations. The model is based on physically motivated scaling laws and simple differential equations describing the shock propagation, the contraction of the neutron star, the neutrino emission, the heating conditions, and the explosion energetics. Using model parameters compatible with multi-D simulations and a fine grid of thousands of supernova progenitors, we obtain a variegated landscape of neutron star and black hole formation similar to other parametrized approaches and find good agreement with semi-empirical measures for the ‘explodability’ of massive stars. Our predicted explosion properties largely conform to observed correlations between the nickel mass and explosion energy. Accounting for the coexistence of outflows and downflows during the explosion phase, we naturally obtain a positive correlation between explosion energy and ejecta mass. These correlations are relatively robust against parameter variations, but our results suggest that there is considerable leeway in parametric models to widen or narrow the mass ranges for black hole and neutron star formation and to scale explosion energies up or down. Our model is currently limited to an all-or-nothing treatment of fallback and there remain some minor discrepancies between model predictions and observational constraints.