Søren-Aksel Sørensen

The Genomic Threading Database: a comprehensive resource for structural annotations of the genomes from key organisms

Currently, the Genomic Threading Database (GTD) contains structural assignments for the proteins encoded within the genomes of nine eukaryotes and 101 prokaryotes. Structural annotations are carried out using a modified version of GenTHREADER, a reliable fold recognition method. The Gen THREADER annotation jobs are distributed across multiple clusters of processors using grid technology and the predictions are deposited in a relational database accessible via a web interface at http://bioinf.cs.ucl.ac.uk/GTD. Using this system, up to 84% of proteins encoded within a genome can be confidently assigned to known folds with 72% of the residues aligned. On average in the GTD, 64% of proteins encoded within a genome are confidently assigned to known folds and 58% of the residues are aligned to structures.

Protocol design for large group multicasting: the message distribution protocol

Abstract Typical multicast protocols are designed to operate with group sizes of up to 50 members. Many of the techniques used do not scale to larger groups very well due to the implosion problem. This paper describes a protocol which has been designed to be scalable to a large degree. The protocol uses negative acknowledgements to increase its reliability, with a saturation method being used when the message is small. This ensures that a message has a high probability of being detected, regardless of the message size. The use of negative acknowledgements is intended to reduce the implosion effect to a minimum. A logical tokenbased ring system is used at the transport level to prevent all of the recipients of a multicast replying at the same time, preventing the client being overrun. Also, this token is used as a failure detector, causing the ring structure to reconfigure around the failure.

Network QoS for Grid Systems

Grid users may wish to have fine-grained control of quality of service (QoS) guarantees in a network in order to allow timely data transfer in a distributed application environment. We present a discussion of the issues and problems involved, with some critical analysis. We propose possible solutions by making reference to and analysing existing work. Also, we describe the mechanisms being proposed as part of a work-in-progress (being conducted by the authors) that uses a peer-to-peer approach to micro-manage network capacity allocations at the edge of the network, at end-sites, in a multi-domain scenario. Scheduling controllers at the end-sites are employed, which are subject to local administrative controls and have flexibility in resource allocation based on user requests for network capacity. We highlight the issues in scaling such systems to large numbers of users and the issues concerning the interfaces available to applications and end-users for accessing such services.

On the dynamics of the Köfels sturzstrom

Geomorphologic data from the Kofels landslide in the Otz valley are compared with the results of computer modelling. The study discusses the dynamic information that can be extracted from the data and argues that a standard Coulomb friction model cannot adequately explain the observed features.

High throughput profile-profile based fold recognition for the entire human proteome

BackgroundIn order to maintain the most comprehensive structural annotation databases we must carry out regular updates for each proteome using the latest profile-profile fold recognition methods. The ability to carry out these updates on demand is necessary to keep pace with the regular updates of sequence and structure databases. Providing the highest quality structural models requires the most intensive profile-profile fold recognition methods running with the very latest available sequence databases and fold libraries. However, running these methods on such a regular basis for every sequenced proteome requires large amounts of processing power.In this paper we describe and benchmark the JYDE (Job Yield Distribution Environment) system, which is a meta-scheduler designed to work above cluster schedulers, such as Sun Grid Engine (SGE) or Condor. We demonstrate the ability of JYDE to distribute the load of genomic-scale fold recognition across multiple independent Grid domains. We use the most recent profile-profile version of our mGenTHREADER software in order to annotate the latest version of the Human proteome against the latest sequence and structure databases in as short a time as possible.ResultsWe show that our JYDE system is able to scale to large numbers of intensive fold recognition jobs running across several independent computer clusters. Using our JYDE system we have been able to annotate 99.9% of the protein sequences within the Human proteome in less than 24 hours, by harnessing over 500 CPUs from 3 independent Grid domains.ConclusionThis study clearly demonstrates the feasibility of carrying out on demand high quality structural annotations for the proteomes of major eukaryotic organisms. Specifically, we have shown that it is now possible to provide complete regular updates of profile-profile based fold recognition models for entire eukaryotic proteomes, through the use of Grid middleware such as JYDE.

The Genomic Threading Database

UNLABELLED The Genomic Threading Database currently contains structural annotations for the genomes of over 100 recently sequenced organisms. Annotations are carried out by using our modified GenTHREADER software and through implementing grid technology. AVAILABILITY http://bioinf.cs.ucl.ac.uk/GTD

Adaptive MANET Routing for Low Overhead

In wireless mission-critical systems, systems may be resource-constrained including limited bandwidth, so minimising protocol overhead, whilst maintaining performance, is important. Proactive MANET routing protocols tend to provide smaller route discovery latency than on-demand protocols because they maintain route information to all the nodes in the network at all time. However, such protocols may impose excessive soft-state routing control overhead which is generated by disseminating periodic update messages. In order to mitigate the side effects of the soft-state control over-heads, we propose two adaptive proactive routing algorithms, namely DT_MIAD and DT_ODPU. By tuning the value of refresh intervals dynamically and automatically, refresh updates are triggered based on traffic conditions and node mobility. We show through simulations that the proposed adaptive routing algorithms can outperform a traditional proactive routing protocol (OLSR).

Analysing the Impact of Topology Update Strategies on the Performance of a Proactive MANET Routing Protocol

This paper presents an analysis of several topology update strategies for proactive MANET routing protocols. Although there have been a number of performance studies of proactive MANET routing protocols, little attention has been paid to the impacts of topology update strategies on routing performance. The goal of this paper is to better understand how topology update strategies can contribute to topology maintenance in proactive mobile ad hoc networks and thus impact the overall performance. Our contribution includes (1) a quantitative analysis on the impacts of proactive update intervals on the routing performance; (2) evaluating the performance of reactive topology updates and proactive updates for proactive routing protocols.

Traffic measurements on a local area computer network

Abstract The characteristics of data traffic on the Ethernet LAN at UCLs Department of Computer Science are investigated on the basis of measurements. The analysis of the arrival pattern of packets on the network is found not to be a Poisson process. Both Weibull and hyperexponential distributions are found to be suitable candidates, with the hyperexponential giving a slightly better fit. Packet arrivals are found to exhibit a source locality similar to that found by Jain and Routhier 1 . It is observed that file transfers are responsible for about 70% of the traffic on the network. Our results will be useful for modelling purposes.

A Cost Ef?cient Framework for Managing Distributed Resources in a Cluster Environment

Most resource management systems in large distributed environments rely on a centralised controller scheme, known as a meta-scheduler. Such a meta-scheduler is designed to have a complete control of all resources worldwide through a number of local schedulers on each participating site. Faced with a potential large number of sites, the central scheduling point represents a potential bottleneck for requests needing fast response. In this paper, we consider the ef¿ciency of a localised two-tier system: the ¿rst tier selects appropriate computing resources from the worldwide pool and creates a virtual cluster. The second tier employs a local scheduler to distribute jobs/tasks from the applications among the resources that make up the cluster. The two-tier approach differentiates the tasks of securing a suitable resource level and managing their use. The aim is to ensure a rapid request response while at the same time optimising the system performance. To this end, we consider a management strategy that incorporates rental policy and scheduling schemes. We propose aggressive and conservative heuristic based policies and evaluate their performance to demonstrate the cost effectiveness of such approaches.