Assembly of viruses that have hundreds of subunits or folding of proteins that have hundreds of amino acids-complex biological reactions-are often spontaneous and rapid. Here, we examine the complete set of intermediates available for the assembly of a hypothetical viruslike particle and the connectivity between these intermediates in a graph-theory-inspired study. Using a build-up procedure, assuming ideal geometry, we enumerated the complete set of 2,423,313 species for formation of an icosahedron from 30 dimeric subunits. Stability of each n-subunit intermediate was defined by the number of contacts between subunits. The probability of forming an intermediate was based on the number of paths to it from its precedecessors. When defining population subsets predicted to have the greatest impact on assembly, both stability- and probability-based criteria select a small group of compact and degenerate species; ergo, only a few hundred intermediates make a measurable contribution to assembly. Though the number of possible intermediates grows combinatorially with the number of subunits in the capsid, the number of intermediates that make a significant contribution to the reaction grows by a much smaller function, a result that may contribute to our understanding of assembly and folding reactions.

No abstract

No abstract

No abstract

Parallel computing is increasingly a requirement of Computational Science and Engineering (CSE) software because problems of interest are very large and hardware systems are becoming parallel through multicore technologies. Inexperienced programmers-non-computer scientists with one semester to a few years of programming experience-are crucial to CSE software development because academic research teams rely on them as application developers. For this group, the basic concepts of parallelism can be explained by analogies rather than through exploring the specifics of various technologies. These analogies are also useful for computer scientists who are new to parallel computing.

No abstract

In fluid flow piping systems, tee and wye junctions are commonly encountered and the study of flow through them has been well documented. Most of these studies have focused on flow characterized as turbulent for which there are nearly constant losses in pressure and kinetic energy in the junctions. Laminar flow has received much less attention since it is not frequently observed in macro scale piping systems where pipe diameters are measured in centimeters. The recent increase in use of micro scale flow devices calls for more research into laminar flow behavior that dictates the design and operation of these devices. This paper documents results from computational fluid dynamics (CFD) simulations of flow in planar tee and wye junctions. The junctions studied consisted of circular pipes with two outlets and one inlet. The angles between the tee and wye junctions were fixed to 180 and 60 degrees, respectively. The inlet pipe diameter was fixed at 50 microns and the outlet pipe diameters were chosen to satisfy the continuity equation constrained to have equal velocities in all pipes. The lengths of the inlet and outlet pipes were varied to achieve fully developed flow within the junction. Following a grid resolution study performed on a sample tee junction, a generalized algorithm was designed and implemented to create three-dimensional models of these junctions subject to the former conditions. In the CFD simulations, Reynolds number was varied in the laminar characterized region between 1 and 2000. The simulations calculated static pressure and velocity magnitude values for a number of planes intersecting the junctions along the inlet and outlet pipes. From these values, pressure and kinetic energy gradients were calculated to estimate the static pressure and kinetic energy at the inlet and outlet pipes of each junction. Finally, these inlet and outlet values were used to calculate the stagnation pressure loss coefficient, which reflects dimensionless losses of pressure and kinetic energy for the junction. These coefficients ranged from 1 to 300 for the tee junction and 1 to 400 for the wye junction over the specified range of Reynolds number. The values were inversely proportional to Reynolds number and curve fits were provided for valid ranges.

Production clusters are a common environment for programming assignments in courses involving High Performance Computing, but they can present challenges, especially in the context of use by courses at institutions remote from these resources. In the context of the Oklahoma Cyberinfrastructure Initiative, the University of Oklahoma has been providing its centrally managed clusters for use by courses statewide. This paper explores mechanisms for using this shared, remote resource for teaching, and advocates in favor of such use.