Malleable applications for scalable high performance computing

Iterative applications are known to run as slow as their slowest computational component. This paper introduces malleability, a new dynamic reconfiguration strategy to overcome this limitation. Malleability is the ability to dynamically change the data size and number of computational entities in an application. Malleability can be used by middleware to autonomously reconfigure an application in response to dynamic changes in resource availability in an architecture-aware manner, allowing applications to optimize the use of multiple processors and diverse memory hierarchies in heterogeneous environments. The modular Internet Operating System (IOS) was extended to reconfigure applications autonomously using malleability. Two different iterative applications were made malleable. The first is used in astronomical modeling, and representative of maximum-likelihood applications was made malleable in the SALSA programming language. The second models the diffusion of heat over a two dimensional object, and is representative of applications such as partial differential equations and some types of distributed simulations. Versions of the heat application were made malleable both in SALSA and MPI. Algorithms for concurrent data redistribution are given for each type of application. Results show that using malleability for reconfiguration is 10 to 100 times faster on the tested environments. The algorithms are also shown to be highly scalable with respect to the quantity of data involved. While previous work has shown the utility of dynamically reconfigurable applications using only computational component migration, malleability is shown to provide up to a 15% speedup over component migration alone on a dynamic cluster environment. This work is part of an ongoing research effort to enable applications to be highly reconfigurable and autonomously modifiable by middleware in order to efficiently utilize distributed environments. Grid computing environments are becoming increasingly heterogeneous and dynamic, placing new demands on applications’ adaptive behavior. This work shows that malleability is a key aspect in enabling effective dynamic reconfiguration of iterative applications in these environments.

Boundary nature induces greater group size and group density in habitat edges: an agent-based model revealed

Apart from the amount and distribution of food resources and predation pressure, we sought to clarify habitat spatial features that affect group size and group density. To simulate animal groups interacting with one another, we made an agent-based model in a computational space. Assuming no predation pressure, we change the resource conditions between the habitat (comprising ‘interior’ and ‘edges’) and the ‘exterior’. The results are as follows: when the group density is high throughout the habitat, large groups tend to frequent the edges (i.e., the boundaries between the interior and exterior) even if food resources are scarce in the exterior. Additionally, when the group density is high and the group size variation is small, both group size and group density increase at the edges. These findings are discussed with reference to primates, particularly the Japanese macaque (Macaca fuscata) as an example of group-living animal.     

On-line feedback-based automatic resource configuration for distributed applications

A key problem in executing performance critical applications on distributed computing environments (e.g. the Grid) is the selection of resources. Research related to “automatic resource selection” aims to allocate resources on behalf of users to optimize the execution performance. However, most of current approaches are based on the static principle (i.e. resource selection is performed prior to execution) and need detailed application-specific information. In the paper, we introduce a novel on-line automatic resource selection approach. This approach is based on a simple control theory: the application continuously reports the Execution Satisfaction Degree (ESD) to the middleware Application Agent (AA), which relies on the reported ESD values to learn the execution behavior and tune the computing environment by adding/replacing/deleting resources during the execution in order to satisfy users’ performance requirements. We introduce two different policies applied to this approach to enable the AA to learn and tune the computing environment: the Utility Classification policy and the Desired Processing Power Estimation (DPPE) policy. Each policy is validated by an iterative application and a non-iterative application to demonstrate that both policies are effective to support most kinds of applications.     

Supporting high‐performance I/O in QoS‐enabled ORB middleware

To be an effective platform for high‐performance distributed applications, off-the-shelf Object Request Broker (ORB) middleware, such as CORBA, must preserve communication-layer quality of service (QoS) properties both vertically (i.e., network interface ↔ application layer) and horizontally (i.e., end-to-end). However, conventional network interfaces, I/O subsystems, and middleware interoperability protocols are not well-suited for applications that possess stringent throughput, latency, and jitter requirements. It is essential, therefore, to develop vertically and horizontally integrated ORB endsystems that can be (1) configured flexibly to support high-performance network interfaces and I/O subsystems and (2) used transparently by performance-sensitive applications. This paper provides three contributions to research on high-performance I/O support for QoS-enabled ORB middleware. First, we outline the key research challenges faced by high-performance ORB endsystem developers. Second, we describe how our real-time I/O (RIO) subsystem and pluggable protocol framework enables ORB endsystems to preserve high-performance network interface QoS up to applications running on off-the-shelf hardware and software. Third, we illustrate empirically how highly optimized ORB middleware can be integrated with real-time I/O subsystem to reduce latency bounds on communication between high-priority clients without unduly penalizing low-priority and best-effort clients. Our results demonstrate how it is possible to develop ORB endsystems that are both highly flexible and highly efficient. This revised version was published online in July 2006 with corrections to the Cover Date.     

A Grid-based solution for management and analysis of microarrays in distributed experiments

Several systems have been presented in the last years in order to manage the complexity of large microarray experiments. Although good results have been achieved, most systems tend to lack in one or more fields. A Grid based approach may provide a shared, standardized and reliable solution for storage and analysis of biological data, in order to maximize the results of experimental efforts. A Grid framework has been therefore adopted due to the necessity of remotely accessing large amounts of distributed data as well as to scale computational performances for terabyte datasets. Two different biological studies have been planned in order to highlight the benefits that can emerge from our Grid based platform. The described environment relies on storage services and computational services provided by the gLite Grid middleware. The Grid environment is also able to exploit the added value of metadata in order to let users better classify and search experiments. A state-of-art Grid portal has been implemented in order to hide the complexity of framework from end users and to make them able to easily access available services and data. The functional architecture of the portal is described. As a first test of the system performances, a gene expression analysis has been performed on a dataset of Affymetrix GeneChip Rat Expression Array RAE230A, from the ArrayExpress database. The sequence of analysis includes three steps: (i) group opening and image set uploading, (ii) normalization, and (iii) model based gene expression (based on PM/MM difference model). Two different Linux versions (sequential and parallel) of the dChip software have been developed to implement the analysis and have been tested on a cluster. From results, it emerges that the parallelization of the analysis process and the execution of parallel jobs on distributed computational resources actually improve the performances. Moreover, the Grid environment have been tested both against the possibility of uploading and accessing distributed datasets through the Grid middleware and against its ability in managing the execution of jobs on distributed computational resources. Results from the Grid test will be discussed in a further paper.     

Marker-assisted rationalisation of genetic resource collections: a case study in flax using AFLPs

Removing redundant germplasm from collections is one of the options for genebanks to increase the efficiency of their genetic resource management. Molecular characterisation of germplasm is thereby becoming more and more important to verify suspected duplication. AFLPs were used to characterise 29 flax accessions of material derived from research activities (hereafter termed ’’breeder’s line”). Based on similar accession names, the breeder’s lines could be classified into three series (’M 25’, ’Ru’ and ’Rm’) that were expected to contain redundancies. In addition, 12 reference cultivars were analysed. A total number of 144 polymorphic bands (59.8%) were scored among the 164 individuals investigated. In general, relatively high levels of intra-accession variation were found, even for the cultivars examined. This finding was not in line with the low outcrossing rates reported for flax. A cluster analysis grouped the ’Ru’ and ’Rm’ series together, indicating their close genetic relationship. An analysis of molecular variance (AMOVA) showed a significant group effect (fibre/oil flax) only for ’M 25’, explaining 34% of the variation observed within this series. For the cultivars 40.5% of the variation was distributed among accessions within groups and all pairwise comparisons were significantly different, except for one case. Both for the series of breeder’s lines and the cultivars the major part of the variation was distributed among individuals within accessions. This component constituted 80.7% and 83.6% of the total variation for the ’Ru’ and ’Rm’ series, respectively. Pairwise comparisons of accessions were performed by AMOVA in order to identify redundant germplasm. Stepwise bulking of accessions until all remaining accessions were significantly different showed that the 29 accessions of breeder’s lines could be reduced to 14. Only a small negative effect of this bulking approach on the among-population component of variance was observed, showing a reduction of 2.6%. Results are discussed in relation to improving the efficiency of collection management. Received: 25 June 2000 / Accepted: 27 October 2000     

An Evaluation of Alternative Designs for a Grid Information Service

Computational grids consisting of large and diverse sets of distributed resources have recently been adopted by organizations such as NASA and the NSF. One key component of a computational grid is an information services that provides information about resources, services, and applications to users and their tools. This information is required to use a computational grid and therefore should be available in a timely and reliable manner. In this work, we describe the Globus information service, describe how this service is used, analyze its current performance, and perform trace-driven simulations to evaluate alternative implementations of this grid information service. We find that the majority of the transactions with the information service are changes to the data maintained by the service. We also find that of the three servers we evaluate, one of the commercial products provides the best performance for our workload and that the response time of the information service was not improved during the single experiment we performed with data distributed across two servers.     

caGrid: design and implementation of the core architecture of the cancer biomedical informatics grid

MOTIVATION: The complexity of cancer is prompting researchers to find new ways to synthesize information from diverse data sources and to carry out coordinated research efforts that span multiple institutions. There is a need for standard applications, common data models, and software infrastructure to enable more efficient access to and sharing of distributed computational resources in cancer research. To address this need the National Cancer Institute (NCI) has initiated a national-scale effort, called the cancer Biomedical Informatics Grid (caBIGtrade mark), to develop a federation of interoperable research information systems. RESULTS: At the heart of the caBIG approach to federated interoperability effort is a Grid middleware infrastructure, called caGrid. In this paper we describe the caGrid framework and its current implementation, caGrid version 0.5. caGrid is a model-driven and service-oriented architecture that synthesizes and extends a number of technologies to provide a standardized framework for the advertising, discovery, and invocation of data and analytical resources. We expect caGrid to greatly facilitate the launch and ongoing management of coordinated cancer research studies involving multiple institutions, to provide the ability to manage and securely share information and analytic resources, and to spur a new generation of research applications that empower researchers to take a more integrative, trans-domain approach to data mining and analysis. AVAILABILITY: The caGrid version 0.5 release can be downloaded from https://cabig.nci.nih.gov/workspaces/Architecture/caGrid/. The operational test bed Grid can be accessed through the client included in the release, or through the caGrid-browser web application http://cagrid-browser.nci.nih.gov.     

Peach: a simple Perl-based system for distributed computation and its application to cryo-EM data processing

A simple distributed processing system named "Peach" was developed to meet the rising computational demands of modern structural biology (and other) laboratories without additional expense by using existing hardware resources more efficiently. A central server distributes jobs to idle workstations in such a way that each computer is used maximally, but without disturbing intermittent interactive users. As compared to other distributed systems, Peach is simple, easy to install, easy to administer, easy to use, scalable, and robust. While it was designed to queue and distribute large numbers of small tasks to participating computers, it can also be used to send single jobs automatically to the fastest currently available computer and/or survey the activity of an entire laboratory's computers. Tests of robustness and scalability are reported, as are three specific electron cryomicroscopy applications where Peach enabled projects that would not otherwise have been feasible without an expensive, dedicated cluster.     

Models@Home: distributed computing in bioinformatics using a screensaver based approach

MOTIVATION: Due to the steadily growing computational demands in bioinformatics and related scientific disciplines, one is forced to make optimal use of the available resources. A straightforward solution is to build a network of idle computers and let each of them work on a small piece of a scientific challenge, as done by Seti@Home (http://setiathome.berkeley.edu), the world's largest distributed computing project. RESULTS: We developed a generally applicable distributed computing solution that uses a screensaver system similar to Seti@Home. The software exploits the coarse-grained nature of typical bioinformatics projects. Three major considerations for the design were: (1) often, many different programs are needed, while the time is lacking to parallelize them. Models@Home can run any program in parallel without modifications to the source code; (2) in contrast to the Seti project, bioinformatics applications are normally more sensitive to lost jobs. Models@Home therefore includes stringent control over job scheduling; (3) to allow use in heterogeneous environments, Linux and Windows based workstations can be combined with dedicated PCs to build a homogeneous cluster. We present three practical applications of Models@Home, running the modeling programs WHAT IF and YASARA on 30 PCs: force field parameterization, molecular dynamics docking, and database maintenance.     

Influence of dendrimers on red blood cells

Dendrimers, highly branched macromolecules with a specific size and shape, provide many exciting opportunities for biomedical applications. However, most dendrimers demonstrate toxic and haemolytic activity because of their positively charged surface. Masking the peripheral cationic groups by coating them with biocompatible molecules is a method to reduce it. It was proven that modified dendrimers can even diminish haemolytic activity of encapsulated drugs. Experiments confirmed that anionic dendrimers are less haemotoxic than cationic ones. Due to the high affinity of dendrimers for serum proteins, presence of these components in an incubation buffer might also influence red blood cell (RBC)-dendrimer interactions and decrease the haemolysis level. Generally, haemotoxicity of dendrimers is concentration-, generation-, and time-dependent. Various changes in the RBCs’ shape in response to interactions with dendrimers have been observed, from echinocytic transformations through cell aggregation to cluster formation, depending on the dendrimer’s type and concentration. Understanding the physical and chemical origins of dendrimers’ influences on RBCs might advance scientists’ ability to construct dendrimers more suitable for medical applications.     

Service-oriented execution model supporting data sharing and adaptive query processing

To deal with the environment’s heterogeneity, information providers usually offer access to their data by publishing Web services in the domain of pervasive computing. Therefore, to support applications that need to combine data from a diverse range of sources, pervasive computing requires a middleware to query multiple Web services. There exist works that have been investigating on generating optimal query plans. We however in this paper propose a query execution model, called PQModel, to optimize the process of query execution over Web Services. In other words, we attempt to improve query efficiency from the aspect of optimizing the execution processing of query plans.     

A Review of Multi-Responsive Membranous Systems for Rate-Modulated Drug Delivery

Membrane technology is broadly applied in the medical field. The ability of membranous systems to effectively control the movement of chemical entities is pivotal to their significant potential for use in both drug delivery and surgical/medical applications. An alteration in the physical properties of a polymer in response to a change in environmental conditions is a behavior that can be utilized to prepare ‘smart’ drug delivery systems. Stimuli-responsive or ‘smart’ polymers are polymers that upon exposure to small changes in the environment undergo rapid changes in their microstructure. A stimulus, such as a change in pH or temperature, thus serves as a trigger for the release of drug from membranous drug delivery systems that are formulated from stimuli-responsive polymers. This article has sought to review the use of stimuli-responsive polymers that have found application in membranous drug delivery systems. Polymers responsive to pH and temperature have been extensively addressed in this review since they are considered the most important stimuli that may be exploited for use in drug delivery, and biomedical applications such as in tissue engineering. In addition, dual-responsive and glucose-responsive membranes have been also addressed as membranes responsive to diverse stimuli.     

An adaptive middleware for supporting time-critical event response

There are many applications where a timely response to an important event is needed. Often such response can require significant computation and possibly communication, and it can be very challenging to complete it within the time-frame the response is needed. At the same time, there could be application-specific flexibility in the computation that may be desired. This paper presents the design, implementation, and evaluation of a middleware that can support such applications. Each of the services in our target applications could have one or more service parameters, which can be modified, within the pre-specified ranges, by the middleware. The middleware enables the time-critical event handling to achieve the maximum benefit, as per the user-defined benefit function, while satisfying the time constraint. Our middleware is also based on the existing Grid infrastructure and Service-Oriented Architecture (SOA) concepts. We have evaluated our middleware and its support for adaptation using a volume rendering application and a Great Lake forecasting application. The evaluation shows that our adaptation is effective, and has a very low overhead.

Exploitation of pepper EST–SSRs and an SSR-based linkage map

As genome and cDNA sequencing projects progress, a tremendous amount of sequence information is becoming publicly available. These sequence resources can be exploited for gene discovery and marker development. Simple sequence repeat (SSR) markers are among the most useful because of their great variability, abundance, and ease of analysis. By in silico analysis of 10,232 non-redundant expressed sequence tags (ESTs) in pepper as a source of SSR markers, 1,201 SSRs were found, corresponding to one SSR in every 3.8 kb of the ESTs. Eighteen percent of the SSR–ESTs were dinucleotide repeats, 66.0% were trinucleotide, 7.7% tetranucleotide, and 8.2% pentanucleotide; AAG (14%) and AG (12.4%) motifs were the most abundant repeat types. Based on the flanking sequences of these 1,201 SSRs, 812 primer pairs that satisfied melting temperature conditions and PCR product sizes were designed. 513 SSRs (63.1%) were successfully amplified and 150 of them (29.2%) showed polymorphism between Capsicum annuum ‘TF68’ and C. chinense ‘Habanero’. Dinucleotide SSRs and EST–SSR markers containing AC-motifs were the most polymorphic. Polymorphism increased with repeat length and repeat number. The polymorphic EST–SSRs were mapped onto the previously generated pepper linkage map, using 107 F₂ individuals from an interspecific cross of TF68 × Habanero. One-hundred and thirtynine EST–SSRs were located on the linkage map in addition to 41 previous SSRs and 63 RFLP markers, forming 14 linkage groups (LGs) and spanning 2,201.5 cM. The EST–SSR markers were distributed over all the LGs. This SSR-based map will be useful as a reference map in Capsicum and should facilitate the use of molecular markers in pepper breeding.Gibum Yi and Je Min Lee equally contributed to this work.     

Parallel morphological/neural processing of hyperspectral images using heterogeneous and homogeneous platforms

The wealth spatial and spectral information available from last-generation Earth observation instruments has introduced extremely high computational requirements in many applications. Most currently available parallel techniques treat remotely sensed data not as images, but as unordered listings of spectral measurements with no spatial arrangement. In thematic classification applications, however, the integration of spatial and spectral information can be greatly beneficial. Although such integrated approaches can be efficiently mapped in homogeneous commodity clusters, low-cost heterogeneous networks of computers (HNOCs) have soon become a standard tool of choice for dealing with the massive amount of image data produced by Earth observation missions. In this paper, we develop a new morphological/neural algorithm for parallel classification of high-dimensional (hyperspectral) remotely sensed image data sets. The algorithm’s accuracy and parallel performance is tested in a variety of homogeneous and heterogeneous computing platforms, using two networks of workstations distributed among different locations, and also a massively parallel Beowulf cluster at NASA’s Goddard Space Flight Center in Maryland.

Habitat use and mating system of the houbara bustard (Chlamydotis undulata undulata) in a semi-desertic area of North Africa: implications for conservation

Studies of the movements and home-ranges of houbara bustards (Chlamydotis undulata undulata) showed sexual and seasonal differences in the use of space, with a polygynous mating system similar to an ‘exploded-lek’ or a ‘resource-defence-polygyny’, that remains undefined. We used the arthropod biomass as an index of the trophic quality of six defined habitats and we radio-tracked 7 females and 13 males to test whether sexual and seasonal variations in habitat use were related to resource availability, and to verify if critical resources for breeding females were monopolised by males. We analysed habitat selection in both sexes separately. We used the habitat type composition of buffer zones around radio-locations to study annual and seasonal habitat selection and to identify preferred habitats, using the chi-square goodness-of-fit test. Habitat use between sexes and between seasons were compared using MANOVA based on log-ratios of habitat proportions. During the year, and in each season, both sexes appeared to be significantly selective for habitats in comparison to their availability. But males avoided esparto grass, while females used all habitats. Habitat use differed between sexes in the breeding season, but not in the non-breeding season. In spring, when food resources were abundant and uniformly distributed in space, males preferred ‘temporarily flooded areas’ and females preferred ‘reg with tall perennials’ that offered both food and cover for brooding. Critical resources were not monopolised by males and the mating system fulfilled the definition of the ‘exploded-lek’. Leks are key sites for reproduction and should be considered as priority areas in further conservation plans.     

Identification of molecular markers associated with linoleic acid desaturation in Brassica napus

 Linolenic acid is a component of canola oil that is readily oxidized, which results in a reduced frying stability and shelf life of the oil. The reduction of linolenic acid in canola seed has therefore been an important breeding objective for many years. The inheritance of linolenic acid concentrations in seed oil is polygenic and is also strongly influenced by the environment. For these reasons, molecular markers are sought to assist in early and reliable selection of desired low linolenic acid genotypes in breeding programmes. Molecular markers associated with low linolenic acid loci were identified in a doubled-haploid population derived from a cross between the Brassica napus lines, ‘Apollo’ (low linolenic)×YN90-1016 (high linolenic) using RAPDs and bulked segregant analysis. A total of 16 markers were distributed over three linkage groups, which individually accounted for 32%, 14% and 5% of the phenotypic variation in linolenic acid content. The rapeseed fad3 gene was mapped near the locus controlling 14% of the variation. The mode of inheritance appeared to be additive, and a QTL analysis showed that collectively the three loci explained 51% of the phenotypic variation within this population. PCR fragments for low linolenic acid ‘Apollo’ alleles (3% linolenic acid) were identified at all three loci. Simultaneous selection for low linolenic acid ‘Apollo’ alleles at each locus resulted in a group of DH lines with 4.0% linolenic acid. The use of these makers in the breeding programme will enhance the breeding of low linolenic acid B. napus cultivars for production in Canada. Received: 23 September 1997 / Accepted: 21 October 1997     

Subpopulation genetic structure of a plant panmictic population of Castanea sequinii as revealed by microsatellite markers

Castanea squinii Dode, an endemic tree widely distributed in China, plays an important role both in chestnut breeding and forest ecosystem function. The spatial genetic structure within and among populations is an important part of the evolutionary and ecological genetic dynamics of natural populations, and can provide insights into effective conservation of genetic resources. In the present study, the spatial genetic structure of a panmictic natural population of C. sequinii in the Dabie Mountain region was investigated using microsatellite markers. Nine prescreened microsatellite loci generated 29–33 alleles each, and were used for spatial autocorrelation analysis. Based on Moran’s I coefficient, a panmictic population of C. sequinii in the Dabie Mountain region was found to be lacking a spatial genetic structure. These results suggest that a high pollen-mediated gene flow among subpopulations counteract genetic drift and/or genetic differentiation and plays an important role in maintaining a random and panmictic population structure in C. sequinii populations. Further, a spatial genetic structure was detected in each subpopulation’s scale (0.228 km), with all three subpopulations showing significant fine-scale structure. The genetic variation was found to be nonrandomly distributed within 61 m in each subpopulation (Moran’s I positive values). Although Moran’s I values varied among the different subpopulations, Moran’s I in all the three subpopulations reached the expected values with an increase in distances, suggesting a generally patchy distribution in the subpopulations. The fine-scale structure seems to reflect restricted seed dispersal and microenvironment selection in C. sequinii. These results have important implications for understanding the evolutionary history and ecological process of the natural population of C. sequinii and provide baseline data for formulating a conservation strategy of Castanea species. __________ Translated from Acta Phytoecologica Sinica, 2006, 30(1): 147–156 [译自: 植物生态学报]