BMC Bioinformatics BioMed Central Review Trends in life science grid from computing grid to

Bio Med Central BMC Bioinformatics

Open Access Review

Trends in life science grid: from computing grid to knowledge grid Akihiko Konagaya*

Address: Genomic Sciences Center, RIKEN, Yokohama, Japan

Email: Akihiko Konagaya*-konagaya@gsc.riken.jp

* Corresponding author

Abstract

Background: Grid computing has great potential to become a standard cyberinfrastructure for

life sciences which often require high-performance computing and large data handling which

exceeds the computing capacity of a single institution.

Results: This survey reviews the latest grid technologies from the viewpoints of computing grid,

data grid and knowledge grid. Computing grid technologies have been matured enough to solve

high-throughput real-world life scientific problems. Data grid technologies are strong candidates

for realizing "resourceome" for bioinformatics. Knowledge grids should be designed not only from

sharing explicit knowledge on computers but also from community formulation for sharing tacit

knowledge among a community.

Conclusion: Extending the concept of grid from computing grid to knowledge grid, it is possible

to make use of a grid as not only sharable computing resources, but also as time and place in which

people work together, create knowledge, and share knowledge and experiences in a community.

Introduction

Bioinformatics applications often require high-perform-ance computing and large data handling which exceeds the computing capacity of a single institution [1]. Sharing of unpublished data is also important in promoting col-laborative research among institutions, as well as sharing of public databases, bioinformatics tools and web services [2-7]. Biological knowledge, such as ontology and meta data, also plays an important role in analysis of experi-mental data and integrating genome-wide OMICS data including genome, transcriptome, proteome, and other types of data [8,9]. Grid computing is a promising infor-mation technology which meets the above requirements, and has great potential to become a standard cyberinfra-structure for life sciences [10,11]. However, many features of it remain to be improved in terms of availability, per-formance and security, to name a few.

This paper reviews the latest grid technologies for life sci-ences mainly from papers published in the proceedings of international conferences: LS-GRID2004 [12], LSGRID2005 [13], LSGRID2006 [14], CCGRID2006 [15] and NETTAB2006 [16].

The grid technologies can be classified into three catego-ries from the viewpoint of application development: com-puting grids, data grids, and knowledge grids. Although the grid is general enough to execute any type of life sci-ence application, the above classification is helpful for

from International Conference in Bioinformatics – InCoB2006

New Dehli, India. 18–20 December 2006

Published: 18 December 2006

BMC Bioinformatics 2006, 7(Suppl 5):S10doi:10.1186/1471-2105-7-S5-S10

? 2006 Konagaya; licensee BioMed Central Ltd

This is an open access article distributed under the terms of the Creative Commons Attribution License (https://www.sodocs.net/doc/dd15446269.html,/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

understanding the pros and cons of grid technologies when they are used for real life science applications.

The organization of this paper is as follows. The section, "Computing grid" introduces computing grid technolo-gies focusing on virtual screening and large-scale sequence matching from the viewpoint of high-throughput com-puting. The next section, "Data grid" focuses on data grid technologies from the viewpoints of service integration, workflow and security when assuming open grid service architecture (OGSA). The "Knowledge grid" section dis-cusses the requirements of knowledge grid technologies when using a grid as a cyberinfrastructure for knowledge creation based on the Nonaka knowledge spiral between explicit knowledge and tacit knowledge. Finally, a sum-mary of the current status and future perspectives of life science grid technologies is presented.

Computing grid

Bioinformatics applications often have to deal with thou-sands of relatively small independent tasks, each of which costs at most seconds or minutes for computation. This type of computation is referred to as "high throughput computing" and is distinguished from "high performance computing", which aims at short turnaround time on large scale computing using parallel processing tech-niques and special purpose computers [17,18]. Although grid computing aims at parallel and distributed computing, like cluster computing, the two differ in net-work latency and robustness. Network latency among institutions is far longer than that in a system area net-work in clusters even if network throughput performance is the same, for example, a giga-bit per second. In addi-tion, the frequency of remote task failures is much higher in grid computing than in cluster computing due to the overhead of remote task invocation and the heterogeneity of computation pools. Therefore, handling of unexpected node termination and network problems is mandatory in grid computing, especially for lengthy execution jobs which take weeks and months of total time. There are two types of high-throughput computing in life sciences: numerical processing, typified by virtual screening, and symbolic processing, typified by sequence matching. High throughput numerical processing

High throughput numerical processing has become popu-lar in bioinformatics due to the emergence of systems biology, which aims at modeling of biological dynamics in molecules, cells, organs and individuals. Huge compu-tational power is necessary for the simulation of molecu-lar folding, molecular docking, and spatiotemporal molecular interaction, and for the kinetic parameter esti-mation of metabolic pathways and signal transduction pathways, and so on. Problem decomposition techniques such as parameter sweep and stochastic modeling are often used to obtain a set of independent tasks in life sci-ence applications.

One of the best examples of life science high-throughput computing is the WISDOM high-throughput docking project in the Enabling Grids for E-sciencE (EGEE) project. It achieved over 46 million docking simulations, using 1700 computers distributed in 15 countries in about 6 weeks. The equivalent of 80 years on a single machine was used to find new inhibitors for a family of proteins pro-duced by Plasmodium falciparum from 11 July 2005 to 19 August 2005 [19].

DIANE is an enhanced version of WISDOM with a light-weight framework. It was used to search for potential drugs for the predicted variants of the avian flu virus (H5N1), and produced two millions docking complexes with a size of 600 gigabytes using 2000 grid worker nodes distributed in 17 countries [20].

The above virtual screening projects revealed the limita-tions and bottlenecks of the current EGEE infrastructure. Overall grid efficiency was reported to be about 50 per-cent, on average. Server license failure, workload manage-ment failure and site failure were major sources of failures with rates of 23, 10 and 9 percent, respectively [21]. This means that much remains to be accomplished in grid middle-ware in improving availability and performance in solving real-life science problems.

Another example of high-throughput computing in bioin-formatics is parameter estimation of ordinary differential equations for the mathematical modeling of metabolic pathways and signal transduction pathways [22]. Genetic algorithms are often used for estimating optimal parame-ter fitting to biological experimental results [23-25]. Genetic algorithms exhibit high degrees of parallelism, since they require multiple trials with various initial con-ditions as well as fitting function evaluation for each indi-vidual on each generation.

"Parameter Mining" is an alternative approach to genetic algorithms for the parameter estimation of mathematical models [26]. It uses two-dimensional geometrical pat-terns representing parameter-parameter dependencies (PPD) in differential equations, obtained by calculating moment parameters, such as area under the curve (AUC), mean residence time (MRT), and variance of residence time (VRT). Each two-dimensional pattern requires 25*21 measurement points to cover (10 to 6)*(10 to 5) parame-ter ranges, and 370 Gigabytes and 71 single cpu days are required for calculation of 256 geometrical patterns with 2,150,400 simulation in total. This CPU and data-inten-sive approach enables more precise mapping of biological

experimental data on appropriate locations in geometrical patterns with a bird's eye view.

High throughput symbolic processing

Sequence analysis, such as homology searches, genome comparisons and genome-wide analyses, are typical examples of time-consuming high-throughput symbolic processing applications in bioinformatics. Although the human genome sequence project has been concluded, there is still strong demand for high-performance sequence analysis due to the emergence of metagenomic projects and human resequencing projects as well as genome sequencing projects on mammalian and other species [27]. Sequencing data are expected to increase more rapidly as high-throughput DNA sequencing tech-nologies become popular and economical.

Unlike numerical processing, bioinformatics symbolic processing often requires large databases such as DNA and protein sequence databases. Sharing and updating of bio-logical databases on the grid are of key importance in high-throughput symbolic processing such as homology searches, genome comparison and genome-wide scan analyses.

Sharing and updating of biological databases

Sharing and updating of biological databases has become more and more difficult and intractable due to the rapid increase in DNA and genome sequence data. Rapid progress of DNA chip technologies also contributes to the expansion of gene expression databases and SNP data-bases. Automatic updating of databases is necessary to decrease the database maintenance costs, especially when the number of replicas becomes large in grid [28]. In the deployment of genome databases on worker nodes, dupli-cated database copying, disk overflow, unexpected shut-down, version management, and file checksum integrity verification are all concerns, as well as parallel and pipe-lined mechanisms for high-throughput data transfer [29]. EGEE also provides a general framework for sharing repli-cas of biological databases represented by logical filena-mes (LFNs) using a replica manager system (RMS). The framework enables execution of bioinformatics applica-tions on computing elements with randomly replicated LNFs on the storage elements of several grid nodes shared by more than 30,000 CPUs in total [30].

The Genome Analysis and Database Update system (GADU) provides an automated, scalable, high-through-put computational workflow engine that executes bioin-formatics tools (BLAST, BLOCKS, PFam, Chisel and InterPro) with public databases (NCBI RefSeq, PIR, Inter-Pro and KEGG) on multiple Grids of different architec-tures and environment, a collective member of more than 18,000 CPUs contributed by more than 60 institutions [31].

Homology search

BLAST is a typical example of high-throughput symbolic processing in homology searches. Many GRID BLAST implementations have been developed and reported [30-35]. The characteristics of Grid Blast are summarized as follows: (1) prestaging of sequence databases to minimize the runtime overhead of transferal of large sequence data-bases, which often reach several Gigabytes in size, (2) databases update which keeps data consistency on the data-grid, (3) dynamic load balancing of query sequences to avoid unexpected slow responses, especially when deal-ing with thousands of query sequences in heterogeneous computation pools including PC-clusters and desktop computers, and (4) assembling of the results from distrib-uted jobs.

Genome comparison

Genome comparison is one of the most promising life sci-ence applications for grid computing. "The computation will be left behind a tidal wave of genomic data, unless an expandable and flexible large scale computing facility is established" described Sugawara, when investigating hor-izontal gene transfer among 354,606 ORFs extracted from more than 100 microbial genomes using 229 CPUs located in five institutions in 2003 [36]. It should be noted the number of pair-wise sequence comparison increases in proportion to the square of the number of genome sequences. Grid is one of feasible information technologies that can provide huge computation power necessary for this purpose.

Genome-wide scan analysis

Genome-wide scan analysis becomes more and more important but time-consuming in nature. Recent disccov-ery of RNA world reveals the importance of finding highly conserved regions in genome sequences for non-coding genes and microRNA binding regions as well as coding-genes and binding factor regions. SNP-based population genetics and copy number analysis on genome sequence variations are also important applications for a life science grid in near future. Gridification of sequence analysis tools are urgent issues to deal with ever–expanding genome sequences [37,38].

Data grid

"We suggest that the full set of bioinformatics resources–the resourceome–should be explicitly characterized and organized." noted Russ Altman in his article [8]. Resour-ceome requires a uniform interface in which all the bioin-formatics databases and application tools can be accessed through web services and workflow systems in a secure fashion. Ontology and/or meta data are also required to

integrate the bioinformatics services. Data Grids based on Open Grid Service Architecture (OGSA) are beginning to satisfy the above requirements, and will be applicable to practical applications including pharmacogenomics and clinical-trials in the near future.

Integration of bioinformatics services

OGSA provides a general framework for sharing of resources among institutions over firewalls based on the Web Service Resource Framework (WSRF). It enables exe-cution of bioinformatics applications and workflows with remote resources through web services in secure fashion. Metadata and ontology play an important role to fill the semantic gap of heterogeneous databases as follows. The Japanese BioGrid project designed application meta-data and data service metadata to fill the semantic gap among gene-protein databases, interaction databases and compound databases necessary for drug-design using GT3 and OGSA-DAI for the implementation of a heterogene-ous database federation [39]. The @neurIST project devel-oped a service-oriented grid infrastructure to integrate public databases, hospital information, private databases, modeling and simulation using Web Service Level Agree-ments (WSLA) for QoS-enabled computer service [40]. The Sealife project aimed at context-based information integration on a semantic web/grid browser which auto-matically links a host of web servers and Web/Grid serv-ices to the Web content being visiting. Text mining and concept mapping techniques were used for bridging the gap between the free text on the current web and the ontology-based mark-up for the semantic web and the grid services [41].

Bioinformatics workflow

Bioinformatics workflow tools are necessary for end-users to make use bioinformatics web/grid services. Taverna is one such example which provides a workflow language and graphical user interface to facilitate the easy building, running and editing of workflows allowing the integra-tion of resources that are published as Web services [42]. However, the quest for resources becomes a very demand-ing and time-consuming activity, so that a dynamic semantic indexing system of bioinformatics services becomes essential [43]. Searching functionally similar bioinformatics workflows is also important for the recy-clable use of bioinformatics workflows [44]. In addition, automatic generation of bioinformatics is possible if bio-informatics ontology that defines input-output data spec-ification and functional specification is established [45].

A workflow management system is also helpful for deploying grid applications because it enables to encapsu-late architectural differences of heterogeneous grid resources from application users [46-48]. Agents society is another approach to integrate insilico experiments, resource discovery and biological system simulation [49]. Secure data access

Many bioinformatics databases are public and freely avail-able, but it is often the case that access to the data needs to be strictly controlled in distributed collaborative research. A secure framework is needed to access clinical data that exists across regional, national and international boundaries for clinical trials and unbiased evaluations of their outcome [50]. Although Public Key Infrastructures (PKI) is the predominant method for enforcing authenti-cation in a grid community, the Virtual Organization for Trials and Epidemiological Studies (VOTES) project adopted the Internet2 Shibboleth technology to allow a "single sign-on" authentication step between the grid/ data servers and the local database resources [35,50,51]. Knowledge grid

Michael Polanyi, a 20th-century philosopher, commented in his book, The Tacit Dimension, that "we should start from the fact that we can know more than we can tell". This means that knowledge which we can represent on computers is just a part of knowledge which we can create, transfer and share among a community.

The Grid can be considered as a kind of "Ba", a Japanese philosophical concept, that conceptualises time and place where people work together and create knowledge [9]. This "ba" can be designed not only for sharing explicit knowledge but also for sharing tacit knowledge among communities and/or virtual organizations [52]. According to the Nonaka knowledge spiral theory [53], knowledge creation requires a cyclic process of knowledge conversion between tacit knowledge and explicit knowl-edge; (1) Socialization (tacit knowledge to tacit knowl-edge), (2) Externalization (tacit knowledge to explicit knowledge) (3) Combination (explicit knowledge to explicit knowledge) and (4) Internalization (explicit knowledge to tacit knowledge). This has significant insights into what it will take to support the realisation of the Grid amongst our scientific community. This frame-work gives a meta-philosophical approach to rationalise the current Grid phenomemon.

Socialization

Socialization is the first step in formulating a community. Grid portals are helpful for attracting those who are inter-ested in some specific field. However, the role of a portal will be limited if it does not allow formulation of user-defined communities. Knowledge grids should provide social communication system-like facilities in which any participant can formulate a new community and can

recruit other participants. Face-to-face meeting or off-site meeting will be also helpful in promoting mutual under-standing in a community.

Externalization

Externalization is the essence of knowledge creation. It is not too much to say that all research activities are a kind of externalization involving publication of research papers as a final result. In this sense, knowledge grid should provide facilities for participants to publish their knowledge in a community. Web-based dynamic contents are one of the promising ways of publication of knowl-edge [54].

Combination

Combination expands knowledge by the sharing of explicit knowledge in a community. Synergy effects can be expected if participants bring together their own knowl-edge. Grid portals [55-57] and application-oriented grids [58-61] play an essential role in this process. Internalization

Internalization is a process of acquiring tacit knowledge by experience. In order to make use of a grid for real world life science problems, a global bioinformatics environ-ment, that is, a problem solving layer for bioinformatics must be developed on a grid. Gridfication of public data-bases and bioinformatics tools are necessary conditions but not sufficient for this. The bioinformatics environ-ment should provide secure facilities to deal with unpub-lished data and customization facilities to develop one's own bioinformatics environment coordinated with global bioinformatics environment.

Conclusion

Computing grid technologies have been matured enough to solve high-throughput real-world life scientific prob-lems like virtual screening of docking simulation. Scalable distributed storage management systems are also neces-sary to deal with high-throughput sequence analysis on ever-increasing DNA sequence data.

Data grid technologies are strong candidate for realizing resourceome for bioinformatics. OGSA and workflow man-agement system enable to develop a global bioinformatics environment in which any biological databases and bio-informatics tools can be access through grid services. Ontology and common data-exchange formats are keys to establish interoperability among bioinformatics grid serv-ices.

Knowledge grid should be designed not only from sharing explicit knowledge on computers but also from commu-nity formulation for sharing tacit knowledge among a community. Then, we can extend the concept of grid as a ba, that is, time and place in which people work together, create knowledge, and share knowledge and experiences in a community.

Acknowledgements

The authors express special thanks for the member of the Open Bioinfor-matics Grid project and anonymous reviewers for their valuable discussion and useful comments for this manuscript.

This article has been published as part of BMC Bioinformatics Volume 7, Sup-plement 5, 2006: APBioNet – Fifth International Conference on Bioinfor-matics (InCoB2006). The full contents of the supplement are available online at https://www.sodocs.net/doc/dd15446269.html,/1471-2105/7?issue=S5. References

1.Krishnan A: A Survey of life sciences applications on the grid.

New Generation Comput 2004, 22:111-126.

2.Li W, Byrnes R, Hayes J, Birnbaum A, Reyes V, Shahab A, Mosley C,

Pekurovsky D, Quinn G, Shindyalov I, Casanova H, Ang L, Berman F, Arzberger P, Miller M, Bourne P: The encyclopedia of life project: grid software and deployment.New Generation Comput 2004, 22:127-136.

3.Hartzwood M, Jirotka M, Procter R, Slack R, Voss A, Lloyd S: Work-

ing IT out in e-Science: Experiences of requirements capture in a HealthGrid project.Proceedings of the HealthGrid2005: Oxford 2005. 7–9 April 2005

4.Seitz L, Montagnat J, Pierson J, Oriol D, Lingrand D: Authentication

and authorization prototype on the micro-grid for medical data management.Proceedings of the HealthGrid2005: Oxford 2005.

7–9 April 2005

5.Zhang N, Rector A, Buchan I, Shi Q, Kalra D, Rogers J, Goble C,

Walker S, Ingram D, Singleton P: A Linkable identity privacy algorithm for HealthGrid.Proceedings of the HealthGrid2005: Oxford 2005. 7–9 April 2005

6.Umetsu R, Ohki S, Fukuzaki A, Konagaya A, Shinbara D, Saito M,

Watanabe K, Kitagawa T, Hoshino T: An Architectural Design of Open Genome Services. In Grid Computing in Life Science (LSGRID2005) Edited by: Tan TW, Arzberger P, Konagaya A. Singapore: World Scientific; 2006:87-98.

7.Konishi F, Yagi T, Konagaya A: MolWorks+G: Integrated Plat-

form for the Acceleration of Molecular Design by Grid Com-puting. In Grid Computing in Life Science (LSGRID2005) Edited by: Tan TW, Arzberger P, Konagaya A. Singapore: World Scientific; 2006:134-141.

8.Cannata N, Merelli E, Altman R: Time to organize the bioinfor-

matics resourceome.PLoS Comput Biol 2005, 1:e76.

9.Konagaya A: OBIGrid: Towards the 'Ba' for sharing resources,

services and knowledge for bioinformatic.Proceedings of the CCGRID2006 BioGrid Workshop: Singapore 2006. 16–19 May 2006 10.Arzberger P, Farazdel A, Konagaya A, Ang L, Shimojo S, Stevens R:

Life sciences and cyberinfrastructure: dual and interacting revolutions that will drive future science.New Generation Com-put 2004, 22:97-110.

11.Konagaya A, Konishi F, Hatakeyama M, Satou K: The superstruc-

ture toward open bioinformatics grid.New Generation Comput 2004, 22:167-176.

12.Konagaya A, Satou K, Eds: Grid computing in life science (LSGRID2004):

Lecture Notes in Bioinformatics LNBI3370Berlin Heidelberg New York: Springer; 2005.

13.Tan T, Arzberger P, Konagaya A, Eds: Grid Computing in Life Science

(LSGRID2005)Singapore: World Scientific; 2006.

14.LSGRID [https://www.sodocs.net/doc/dd15446269.html,/]

https://www.sodocs.net/doc/dd15446269.html,GRID2006 [https://www.sodocs.net/doc/dd15446269.html,.sg/ccgrid2006/]

https://www.sodocs.net/doc/dd15446269.html,TAB [https://www.sodocs.net/doc/dd15446269.html,/]

17.Taiji M, Narumi T, Ohno Y, Futatsugi N, Suenaga A, Takada N, Kon-

agaya A: Protein Explorer: A Petaflops Special-Purpose Com-puter System for MolecularDynamics Simulations.

Proceedings of the Supercomputing 2003 in CD-ROM 2003.

18.Masuno S, Maruyama T, Yamaguchi Y, Konagaya A: Multidimen-

sional Dynamic Programming for Homology Search on Dis-tributed Systems.Proceedings of European Conference on Parallel Computing (Euro-Par2006): September 2006; Dresden 2006:1127-1137.

19.Breton V, Kasam V, Jacq N: High Throughput Grid Enabled Vir-

tual Screening.Proceedings of the NETTAB2006: Santa Margherita 2006:14-18. 10–13 July 2006

20.Lee H, Salzemann J, Jacq N, Ho L, Chen H, Breton V, Merelli L,

Milanesi L, Lin S, Wu Y: Grid-enabled High Throughput in-silico Screening Against Influenza A Neuraminidase.Proceedings of the NET-TAB2006: Santa Margherita 2006:19-25. 10–13 July 2006 21.Jacq N, Breton B, Chen H, Ho L, Hofmann M, Lee H, Legre Y, Lin S,

Maas A, Medernach E, Merelli I, Milanesi L, Rastelli G, Reichstadt M, Salzemann J, Schwichtenberg H, Sridhar M, Kasam V, Wu Y, Zimmer-mann M: Large Scale In Silico Screening on Grid Infrastruc-tures.Proceedings of the LSGRID2006: Yokohama 2006:123-136. 13–

14 October 2006

22.Sugimoto M, Takahashi K, Kitayama T, Ito D, Tomita M: Distributed

Cell Biology Simulations with E-Cell System. In Grid Computing in Life Science (LS-GRID2004) Edited by: Konagaya A, Satou K. Berlin Hei-delberg New York: Springer; 2005:20-31. [Lecture Notes in Bioinfor-matics, vol 3370]

23.Kimura S, Kawasaki T, Hatakeyama M, Naka T, Konishi F, Konagaya

A: OBIYagns: a grid-based biochemical simulator with a parameter estimator.Bioinformatics 2004, 20:1646-1648.

24.Imade H, Mizuguchi N, Ono I, Ono N, Okamoto M: Gridifying an

Evolutionary Algorithm for Inference of Genetic Networks U sing the Improved GOGA Framework and Its Perform-ance Evaluation on OBI Grid. In Grid Computing in Life Science (LSGRID2004) Edited by: Konagaya A, Satou K. Berlin Heidelberg New York: Springer; 2005:171-186. [Lecture Notes in Bioinformatics, vol 3370]

25.Kimura S, Ide K, Kashihara A, Kano M, Hatakeyama M, Masui R, Nak-

agawa N, Yokoyama S, Kuramitsu S, Konagaya A: Inference of S-system models of genetic networks using a cooperative coevolutionary algorithm.Bioinformatics 2005, 21:1154-1163. 26.Konagaya A, Azuma R, Umetsu R, Ohki S, Konishi F, Matsumura K,

Yoshikawa S: Parameter Mining: Discovery of Dynamical Characteristics using Geometrical Patterns of Parameter-Parameter Dependencies on Differential Equations.Proceed-ings of the LSGRID2006: Yokohama :137-152. 13–14 October 2006

27.Ensembl [https://www.sodocs.net/doc/dd15446269.html,/]

28.Salzemann J, Jacq N, Le Mahec G, reton V: Replication and Update

of Molecular Biology Databases in a Grid Environment.Pro-ceedings of the NET-TAB2006: Santa Margherita 2006:33-37. 10–13 July 2006

29.Satou K, Tsuji S, Nakashima Y, Konagaya A: Parallel and Pipelined

Database Transfer in a Grid Enviroment for Bioinformatics.

In Grid Computing in Life Science (LSGRID2005) Edited by: Tan TW, Arz-berger P, Konagaya A. Singapore: World Scientific; 2006:32-49.

30.Blanchet C, Combet C, Deleag G: Integrating Bioinformatics

Resources on the EGEE Grid Platform.Proceedings of the CCGRID2006 BioGrid Workshop: Singapore 2006. 16–19 May 2006 31.Sulakhe D, Rodriguez A, Wilde M, Foster I, Maltsev N: Using multi-

ple Grid resources for Bioinformatics applications in GADU.

Proceedings of the CC-GRID2006 BioGrid Workshop: Singapore 2006. 16–

19 May 2006

32.Krishnan A: GridBLAST: a Globus-based high-throughput

implementation of BLAST in a Grid computing framework, Concurrency and Computation.Practice and Experience 2004, 17:1607-1623.

33.Satou K, Nakashima Y, Tsuji S, Defago X, Konagaya A: An Inte-

grated System for Distributed Bioinformatics Environment on Grids. In Grid Computing in Life Science (LSGRID2004) Edited by: Konagaya A, Satou K. Berlin Heidelberg New York: Springer; 2005:8-19.

[Lecture Notes in Bioinformatics, vol 3370]

34.Konishi F, Konagaya A: The Architectural Design of High-

Throughput BLAST Services on OBIGrid. In Grid Computing inLife Science (LSGRID2004) Edited by: Konagaya A, Satou K. Berlin Hei-delberg New York: Springer; 2005:32-42. [Lecture Notes in Bioinfor-matics, vol 3370]

35.Sinnott R, Ajayi O, Stell A, Jiang J, Watt J: User-Oriented Access

to Secure Biomedical Resources through the Grid.Proceed-ings of the LSGRID2006: Yokohama 2006:71-86. 13–14 October 2006 36.Sugawara H: Gene Trek in Procaryote Space Powered by a

GRID Environment. In Grid Computing in Life Science (LSGRID2004) Edited by: Konagaya A, Satou K. Berlin Heidelberg New York: Springer;

2005:1-7. [Lecture Notes in Bioinformatics, vol 3370]37.Loong S, Mishra S: Gridifying Viral MicroRNAs Identification.

Proceedings of the LSGRID2006: Yokohama 2006:7-24. 13–14 October 2006

38.Rajapakse J, Chen C: Computational grid for comparative

genomics to identify conserved non-coding regions.Proceed-ings of the LSGRID2006: Yokohama 2006:25-36. 13–14 October 2006 39.Tohsato Y, Kosaka T, Date S, Shimojo S, Matsuda H: Heterogene-

ous Database Federation using Grid Technology for Drug Discovery Process. In Grid Computing in Life Science (LSGRID2004) Edited by: Konagaya A, Satou K. Berlin Heidelberg New York: Springer;

2005:43-52. [Lecture Notes in Bioinformatics, vol 3370]

40.Arbona T, Benkner S, Fingberg J, Engelbrecht G, Hof-mann M, Kumpf

K, Lonsdale G, Woehrer A: A Service-oriented Grid Infrastruc-ture for Biomedical Data and Compute Services.Proceedings of the NETTAB2006: Santa Margherita 2006:50-54. 10–13 July 2006 41.Schroeder M, Burger A, Kostlova P, Stevens R, Haber-mann B, Dieng-

Kuntz R: From a Services-based eScience Infrastructure to a Semantic Web for the Life Sciences: The Sealife Project.Pro-ceedings of the NETTAB2006: Santa Margherita 2006:26-30. 10–13 July 2006

42.Oinn T, Addis M, Ferris J, Marvin D, Senger M, Greenwood M, Carver

T, Glover K, Pocock M, Wipat A, Li P: Taverna: a tool for the composition and enactment of bioinformatics workflows.

Bioinformatics 2004, 20:3045-3054.

43.Falzone A, Melato M, Porro I, Ratto S, Schenone A, Torterolo L: A

GRID-based multilayer architecture for bioinformatics.Pro-ceedings of the NETTAB2006: Santa Margherita 2006:45-49. 10–13 July 2006

44.Seo J, Senoo S, Takenaka Y, Matsuda H: Extraction of Functionally

Similar Bioinformatics Workflows.Proceedings of the NETTAB2006: Santa Margherita 2006:70-74. 10–13 July 2006

45.Konagaya A: Bioinformatics Ontology: Towards the Automat-

ics Generation of Bioinformatics Workflow for Web Serv-ices.Proceedings of the NETTAB2006: Santa Margherita 2006:75-82.

10–13 July 2006

46.Birnbaum A, Hayes J, Li W, Miller M, Arzberger P, Bourne P, Casa-

nova H: Grid Workflow Software for a High-Throughput Pro-teome Annotation Pipeline. In Grid Computing in Life Science (LSGRID2004) Edited by: Konagaya A, Satou K. Berlin Heidelberg New York: Springer; 2005:68-81. [Lecture Notes in Bioinformatics, vol 3370]

47.Pan M, Toga A: A grid enabled workflow management system

for managing parameter sweep applications in neuroimag-ing research.Proceedings of the CCGRID2006 BioGrid Workshop: Sin-gapore 2006. 16–19 May 2006

48.Shimosaka H, Hiroyasu T, Miki M: Distributed Workflow Man-

agement System based on Publish-Subscribe Notification for Web Services.Proceedings of the LSGRID2006: Yokohama 2006:93-105. 13–14 October 2006

49.Bartocci E, Cacciagrano D, Cannata N, Corradini F, Merelli E, Milanesi

L, Romano P: A Grid infrastructure for managing workflows in bioinformatics applications.Proceedings of the NETTAB2006: Santa Margherita 2006:38-44. 10–13 July 2006

50.Stell A, Sinnott R, Ajayi O: Secure, Reliable and Dynamic Access

to Distributed Clinical Data.Proceedings of the LSGRID2006: Yoko-hama 2006:56-70. 13–14 October 2006

51.Sinnott R, Bayliss C: Towards Data Grids for Microarray

Expression Profiles.Proceedings of the LSGRID2006: Yokohama 2006:37-55. 13–14 October 2006

52.Konagaya A: Grid as a "Ba" for Biomedical Knowledge Crea-

tion. In Grid Computing in Life Science (LS-GRID2005) Edited by: Tan T, Arzberger P, Konagaya A. Singapore: World Scientific; 2006:1-10.

53.Nonaka I, Toyama R, Konno N: SECI, Ba and leadership: a uni-

fied model of dynamic knowledge creation.Long Range Planning 2000, 33:5-34.

54.Konishi F, Ishii M, Ohki S, Umetsu R, Konagaya A: RABC: New Bar-

rier-less Approach for Public Computing Platform.Proceed-ings of the LS-GRID2006: Yokohama 2006:106-116. 13–14 October 2006

55.Shahab A, Chuon D, Suzumura T, Li W, Byrnes R, Tanaka K, Ang L,

Matsuoka S, Bourne P, Miller M, Arzberger P: Grid Portal Interface for Interactive Use and Monitoring of High-Throughput Pro-teome Annotation. In Grid Computing in Life Science (LSGRID2004) Edited by: Konagaya A, Satou K. Berlin Heidelberg New York: Springer;

2005:53-67. [Lecture Notes in Bioinformatics, vol 3370]

56.Li W: Building cyberinfrastructure for bioinformatics using

service oriented architecture.Proceedings of the CCGRID2006 BioGrid Workshop: Singapore 2006. 16–19 May 2006

57.Fukuzaki A, Nagashima T, Ide K, Konishi F, Hatakeyama M, Yokoyama

S, Kuramitsu S, Konagaya A: Genome-Wide Functional Annota-tion Environment for em Thermus Thermophilus in OBIGrid. In Grid Computing in Life Science (LSGRID2004) Edited by: Konagaya A, Satou K. Berlin Heidelberg New York: Springer; 2005. [Lec-ture Notes in Bioinformatics, vol 3370]

58.DAscia S, Frangiamone G: Clinical-Genomics data modelling

using HL7 standards in GebbaLab project.Proceedings of the NETTAB2006: Santa Margherita 2006:109-117. 10–13 July 2006

59.Fato M, Papadimitropoulos A, Porro I, Scaglione S, Schenone A,

Torterolo L, Viti F: A Grid Approach for Large Data Processing in Biomedicine.Proceedings of the NETTAB2006: Santa Margherita 2006:118-123. 10–13 July 2006

60.Emerson A, Rossi E: ImmunoGrid – The virtual human immune

system project.Proceedings of the NETTAB2006: Santa Margherita 2006:124-128. 10–13 July 2006

61.Jones A, White R, Gray W, Bisby F, Caithness N, Pittas N, Xu X, Sut-

ton T, Fiddian N, Culham A, Scoble M, Williams P, Bromley O, Brewer P, Yesson C, Bhagwat S: Building a Biodiversity GRID. In Grid Computing in Life Science (LSGRID2004) Edited by: Konagaya A, Satou K. Berlin Heidelberg New York: Springer; 2005:140-151. [Lecture Notes in Bioinformatics, vol 3370]