Steven J. Johnston

BioSimGrid: grid-enabled biomolecular simulation data storage and analysis

In computational biomolecular research, large amounts of simulation data are generated to capture the motion of proteins. These massive simulation data can be analysed in a number of ways to reveal the biochemical properties of the proteins. However, the legacy way of storing these data (usually in the laboratory where the simulations have been run) often hinders a wider sharing and easier cross-comparison of simulation results. The data is commonly encoded in a way specific to the simulation package that produced the data and can only be analysed with tools developed specifically for that simulation package. The BioSimGrid platform seeks to provide a solution to these challenges by exploiting the potential of the Grid in facilitating data sharing. By using BioSimGrid either in a scripting or web environment, users can deposit their data and reuse it for analysis. BioSimGrid tools manage the multiple storage locations transparently to the users and provide a set of retrieval and analysis tools for processing the data in a convenient and efficient manner. This paper details the usage and implementation of BioSimGrid using a combination of commercial databases, the Storage Resource Broker and Python scripts, gluing the building blocks together. It introduces a case study of how BioSimGrid can be used for better storage, retrieval and analysis of biomolecular simulation data.

Grid computing and biomolecular simulation

Biomolecular computer simulations are now widely used not only in an academic setting to understand the fundamental role of molecular dynamics on biological function, but also in the industrial context to assist in drug design. In this paper, two applications of Grid computing to this area will be outlined. The first, involving the coupling of distributed computing resources to dedicated Beowulf clusters, is targeted at simulating protein conformational change using the Replica Exchange methodology. In the second, the rationale and design of a database of biomolecular simulation trajectories is described. Both applications illustrate the increasingly important role modern computational methods are playing in the life sciences.

Quality assurance for biomolecular simulations

Contemporary structural biology has an increased emphasis on high-throughput methods. Biomolecular simulations can add value to structural biology via the provision of dynamic information. However, at present there are no agreed measures for the quality of biomolecular simulation data. In this Letter, we suggest suitable measures for the quality assurance of molecular dynamics simulations of biomolecules. These measures are designed to be simple, fast, and general. Reporting of these measures in simulation papers should become an expected practice, analogous to the reporting of comparable quality measures in protein crystallography. We wish to solicit views and suggestions from the simulation community on methods to obtain reliability measures from molecular-dynamics trajectories. In a database which provides access to previously obtained simulations [Formula: see text] for example BioSimGrid ( http://www.biosimgrid.org/ ) [Formula: see text] the user needs to be confident that the simulation trajectory is suitable for further investigation. This can be provided by the simulation quality measures which a user would examine prior to more extensive analyses.

.NET gadgeteer: a new platform for K-12 computer science education

In this paper we present the features of a new physical device prototyping platform called Microsoft .NET Gadgeteer along with our initial experiences using it to teach computer science in high schools. Gadgeteer makes it easy for newcomers to electronics and computing to plug together modules with varied functionality and to program the resulting systems behavior. We believe the platform is particularly suited to teaching modern programming concepts such as object-oriented, event-based programming and it could be a timely addition to established teaching tools given the current interest in improving high school computer science education in some regions. We have run a number of pilot studies in the US and in the UK with students of varying age and ability. Our results indicate that the tangible and expressive nature of Gadgeteer helps to engage and motivate a diverse set of students. We were also pleasantly surprised by the level of polish and sophistication of the devices which were built. We hope to further explore the potential of Gadgeteer for teaching in future work and we encourage others to build on our experiences.

Clouds in Space: Scientific Computing using Windows Azure

In this paper we report upon the cloud-based solution that we designed and implemented for space situational awareness. We begin by introducing the background to the work and to the area of space situational awareness. This concerns tracking the hundreds of thousands of known objects in near-Earth orbits, and determining where it is necessary for satellite operators to conduct collision-avoidance manoeuvres to protect their satellites. We also discuss active debris removal, which would be necessary to stabilise the debris population at current levels. We examine the strengths that cloud-based solutions offer in general and how these specifically fit to the challenges of space situational awareness, before describing the architecture we designed for this problem. We demonstrate the feasibility of solving the space situational awareness problem with a cloud-based architecture and note that as time goes on and debris levels rise due to future collisions, the inherent scalability offered by a cloud-based solution will be invaluable.

Exploiting Cloud Computing for Algorithm Development

We consider the application of cloud computing to the process of algorithm development. We introduce a case study focusing on the development of a novel algorithm in computational electromagnetics, illustrating several challenging areas for algorithm developers where cloud-based architectures can deliver enhanced productivity and potentially save costs. The development, verification and tuning of our algorithm have all been assisted by cloud-based technologies. Our preliminary results both demonstrate the potential of the algorithm to solve the problems accurately, and of cloud-based architectures to accelerate the development and verification process. We propose that cloud-based architectures will in the future play a greater role in the development of algorithms; saving costs by improving hardware utilisation, and reducing turnaround time.

Using the MEAN stack to implement a RESTful service for an Internet of Things application

This paper examines the components of the MEAN development stack (MongoDb, Express.js, Angular.js, & Node.js), and demonstrate their benefits and appropriateness to be used in implementing RESTful web-service APIs for Internet of Things (IoT) appliances. In particular, we show an end-to-end example of this stack and discuss in detail the various components required. The paper also describes an approach to establishing a secure mechanism for communicating with IoT devices, using pull-communications.

Scientific Computation and Data Management Using Microsoft Windows Azure

Cloud computing is the next stage in the evolution of computational and data handling infrastructure, establishing scale out from clients, to clusters to clouds. With the use of a case study, Microsoft Windows Azure has been applied to Space Situational Awareness (SSA) creating a system that is robust and scalable, demonstrating how to harness the capabilities of cloud computing. The generic aspects of cloud computing are discussed throughout.

Applicability of commodity, low cost, single board computers for Internet of Things devices

We can expect the number of Internet of Things (IoT) devices to rapidly increase, in part due to the availability of low cost powerful hardware. These advances in hardware introduce a new class of computer — the commodity low-cost Single Board Computer (SBC); for example the Raspberry Pi. These devices are capable of running a full operating system and are accessible to non-technical users. In this paper we demonstrate the feasibility of using this class of computer to construct IoT devices without deep technical knowledge of embedded systems. Our IoT device is targeted at real world data collection and is the basis of a deployed device. We discover that although the existing SBC hardware on the market varies dramatically, much of it is very applicable to IoT scenarios. Generally their default out of the box configuration supports one or more Linux variants, high-level programming languages and standard hardware libraries. They are very configurable and attractively priced — we predict a growth in using this class of computer particularly for IoT prototypes; disposable compute or low hardware volume scenarios. In the future we can expect hardware improvements and enhanced operating systems taking into account the issues surrounding IoT devices, such as SD card storage corruptions, power consumption and improved failure detection.

Managing Large Volumes of Distributed Scientific Data

The ability to store large volumes of data is increasing faster than processing power. Some existing data management methods often result in data loss, inaccessibility or repetition of scientific simulations. We propose a framework which promotes collaboration and simplifies data management. We propose an implementation independent framework to promote collaboration and data management across a distributed environment. The framework features are demonstrated using a .NET Framework implementation called the Storage and Processing Framework.