MOWSERV 2: Friendly and extensible web platform ...

4 downloads 150 Views 184KB Size Report
We present a new integration platform that solves all those issues providing a flexible and friendly environment. Keywords: Integrated Bioinformatics tools, ...
MOWSERV 2: Friendly and extensible web platform for bioinformatics tools integration. Jose Manuel Mateos1, Alfredo Martinez1, Oswaldo Trelles1 1

Computer Architecture Department, University of Malaga, 29071 Málaga, Spain {jmateos, amatrinezl,ots}uma.es

Abstract. In recent years the growth in diversity and heterogeneity of Bioinformatics tools is a fact. The variety of these tools is one of the keys of data analysis flexibility. However, it also carries some disadvantages such as un-standardized service interfaces, different data formats for inputs and outputs, etc. As a direct result, some of the applications that integrate those tools have become outdated for todays needs. We present a new integration platform that solves all those issues providing a flexible and friendly environment. Keywords: Integrated Bioinformatics tools, configurable client.

1

Introduction

MOWServ[1] the current platform offered by the Spanish National Institute of Bioinformatics (INB), provides an integrated access to databases and analytical tools. The main advantages performed by the platform are reliable service invocation, ensured by the scheduler module, asynchronous calls, user data management into a “local” file repository and data type and tool search. MOWServ was developed in 2005 and since then, new technological advances and new community needs have been raised and new Web services definitions and repositories have been created. The user feedback and experience in the last five years has provided us with new user needs knowledge, such as different repositories support, platform configuration dependent on different user skills and support for the constant growth on file size used on bioinformatics tools. In order to extend the MOWServ capabilities, we have developed a new version of the platform, MOWServ2. The new platform can be configured in order to provide specific functionalities or even to develop new components that cover future needs. Therefore, it can be adapted to different types of bioinformatics services, not only BioMoby[2] based. Taking advantage of new technologies in Web developing, we have take special care with system friendliness, offering a more natural and intuitive user interface without sacrificing the system functionality.

MOWSERV 2: Friendly and extensible web platform for bioinformatics tools integration. 2

As a platform independent application, MOWServ2 can be accessed through any common Web navigator, which means no installation is required. The based interface allows the user to load any necessary small program which is also called a component. These components are independent from each other, keeping the system modularity.

2

Architecture

MOWServ2 is based on client-server architecture, the communication of the two parts has been implemented on Web Service. This architecture provides a robust system and it is easily expandable. The Web client which is based on the RIA (Rich Internet Applications) has been implemented. It is a web application that has many of the characteristics of desktop applications. We decided to implement the client side using one of the most extended technologies, Adobe Flex, which compared to similar technologies is Web browser independent. The client is based on small independent applications called components. The client core provides each component a list of all available resources, so if, for example, a component requests tool execution, the system offers a specific component keeping the components independent from each other. The components can work together, sending data and messages between them, this communication improves functionality and allows more powerful data analysis. On the other hand, the server side has been implemented using a Web Service Interface over the mAPI[3], a framework for the development of applications. This allows connection to different data sources and combine its contents. The main functionality served by the mAPI includes multiple repository support, user data management, tools execution etc.

Fig1. MOWServ2 architecture. The top layer represents the platform client side which

is the core of MOWServ2 system. The bottom layer represents the mAPI Web service which provides all the mAPI modules functionalities.

MOWSERV 2: Friendly and extensible web platform for bioinformatics tools integration. 3

A developing framework also allows an easy way to create new components and import them into the platform without modifying it internally.

3

Functionality

MOWServ 2 has been developed as a plug-in based system, so specific or additional functionality can be extended by adding new components. MOWServ 2 current version (on development) includes the following functionality: 3.1

Tools and Data types discovering

Sometimes the amount of tools or data types within a repository can be an issue, especially on large ones. Navigating through the entire tool/data type list searching for a specific one can be a tedious and inefficient work. MOWServ 2 provides a specific component built over Magallanes[4] software library that allows ─ Data types and tools search: the component search into all repository resources (tool or data type) by names, descriptions and also in tool inputs and outputs. ─ Feedback system: the component uses the feedback system implemented on Magallanes as a way to order search results. The order is based on users past selections, so more relevant results will be on the top of the result list. The search results can be used to invoke other components, for example, to create new user data from a specific data type or invoke a specific tool. 3.2

Service invocation

Service invocation and execution is indispensable for any integrated bioinformatics client. MOWServ 2 integrates service execution with the availability of service input and output edition. The execution component has been built using the mAPI deployed web service, which means that service input and outputs can be on different format than the one accepted by the service. 3.3

User data creation and edition

MOWServ2 allows the user having he’s own “local” file repository. The files can be organized into folders, be deleted, renamed etc. Those files can be used as a service input, and they can also be edited. Built over the user data standardization and management library BioData-SF[5], the object edition component allows user data creation and edition, automatic data type interface generation, user data automatic recognition (ongoing work) and on the fly service input creation and output edition.

MOWSERV 2: Friendly and extensible web platform for bioinformatics tools integration. 4

3.4

Additional functionality

Several auxiliary components have been developed as a way to perform general system tasks such as: ─ Data type tree: shows a list of all available data types within a repository. ─ Tool tree: same as previous but showing tool list. ─ File upload: this component allows the users to upload to the system their own files. The component can upload multiple files at the same time. ─ Login manager: this component manages the user login and logout. ─ HTML loader: HTML pages can be loaded into the system using this component.

4

Conclusions

As an approach toward usability and intuitive interfaces, Bioinformatics applications and tool integration are mandatory. Combining integration and a highly flexible platform MOWServ2 client has been developed, providing a configurable platform and an easy way to extend functionality by developing new components.

5

Acknowledgements

This work has been partially financed by the National Institute for Bioinformatics (www.inab.org) a platform of Genoma-España, ACGT-EU project "Advancing Clinico-Genomic Trials on Cancer" (EU-contract no.026996) and RIRAAF: Research network of allergies and drugs adverse reactions. The authors would like to thank the Bitlab research team for invaluable help.

6

References

1. Navas I, Rojano M, Ramirez S, Pérez AJ, Aldana JF, Trelles O: Intelligent client for integrating bioinformatics services. Bioinformatics 2006 , 22:106-111. 2. BioMoby Dashboard. http://biomoby.open-bio.org/CVS_CONTENT/mobylive/Java/docs/Dashboard.html. 3. Ramirez, et al. A flexible framework for the design of knowledge-discovery clients. In: International Conference on Telecommunications and Multimedia. (2008) Ierapetra, Crete, Greece. Available at: http://chirimoyo.ac.uma.es/mapi/documents/Temu-08Ramirez_et_al.pdf. 4. Ríos J, et al. Magallanes: a web services discovery and automatic workflow composition tool. BMC Bioinformatics (2009) 10:334 5. Martinez A, et al. Towards closing the gap between user data and standardized input. In : Network Tools and Applications in Biology(NETTAB (2009) Catalina, Sicily.

Suggest Documents