A central challenge of bioinformatics is the integration of diverse datasets from both public and proprietary biological data sets, and the follow-on construction of software applications that leverage the resulting integrated content that are easily accessible throughout the enterprise. Bioinformatics groups looking to develop a new internal application face a number of commonly recurring challenges including:

• Parsing, formatting, and loading of data from heterogeneous data sources.
• Developing the administrative tools for managing the automatic update of each data source in a timely manner.
• Enabling flexible querying of the integrated data for scientific users wishing to pose a wide variety of research questions.
• Ensuring scalability and technical support of the application to provide for future functional, analytical, or data-content enhancements.

Figure 1. Tracing connections to ask a research question.

The development of integrated biological databases is motivated by the need for scientists to ask deeper scientific questions. For example, the question: "What compounds show activity against targets associated with pathway X" can be investigated by tracing connections from the pathway to its associated genes, the bioassays that target those genes, and the compounds that show activity when screened against those assays (Figure 1).

BioLink is a platform for facilitating the integration of biological data, for updating the data automatically, and for building consistent and powerful web-based interfaces that enable the integrated data content to be searched, explored, extracted, and analyzed (Figure 2). It provides an end-to-end solution for the rapid delivery of biological databases. It includes a database layer for containing both public and proprietary datasets along with meta-data descriptors that define the content and how it should be rendered, a JSP-based service layer for extracting data content in response to user queries, and a user-interface layer built upon a growing collection of powerful and re-usable container and charting widgets. These widgets can be readily organized into pages (perspectives) and separate tabs as required.

Figure 2. Architecture Overview.

Web UI Data Services Integrated Data

A toolkit of reusable Web 2.0 widgets for displaying structured content. Pages are organized into customizable perspectives, each consisting of one or more tabs. Each tab has a distinct layout which contains one or more widgets. JSP framework for serving up annotations (JSON structured content), meta-data (detailed description of the content), and UI definitions (mappings between Web UI widgets and metadata). Integrated data content, including data update services, and an API that serves up both public and proprietary data content as requested by asynchronous data services.

For example, our DataTable widget (Figure 3) is a convenient container for the display of tabular data in any context throughout the interface. It automatically provides support for configuring column layout, multi-column sorting, data export, field-specific URLs, row selection, and a menu of actions that can be triggered against selected items.

Figure 3. DataTable widget displaying the gene results of a search for gpcr*, sorted by Gene Symbol, with several highlighted genes.

Our charting widgets (Figure 4) leverage the Google charting API to enable insertion of both static and dynamically constructed histograms, pie charts, scatter plots, and other visualizations. They can be used in the context of both for data analysis and system administration.

Figure 4. 3d Pie Chart and Bar Chart with IQR, Median, and Mean.

Our BioLink Explorer system is a publicly accessible application that exemplifies some of the key features of the full Enterprise platform. It integrates knowledge of human genes, biological pathways and interactions, disease associations, functional annotations, PubChem bioassays, references, and clinical trials. BioLink Enterprise is flexible enough to be the basis for a wide range of integrated data-centric applications where ease-of-use, scalability, maintainability, and cost-effectiveness are key project objectives.

Please contact us for more information about how BioLink can address your data integration and application development needs.