DraGnET: software for storing, managing and analyzing annotated draft genome sequence data

Abstract

Background: New "next generation" DNA sequencing technologies offer individual researchers the ability to rapidly generate large amounts of genome sequence data at dramatically reduced costs. As a result, a need has arisen for new software tools for storage, management and analysis of genome sequence data. Although bioinformatic tools are available for the analysis and management of genome sequences, limitations still remain. For example, restrictions on the submission of data and use of these tools may be imposed, thereby making them unsuitable for sequencing projects that need to remain in-house or proprietary during their initial stages. Furthermore, the availability and use of next generation sequencing in industrial, governmental and academic environments requires biologist to have access to computational support for the curation and analysis of the data generated; however, this type of support is not always immediately available.

Results: To address these limitations, we have developed DraGnET (Draft Genome Evaluation Tool). DraGnET is an open source web application which allows researchers, with no experience in programming and database management, to setup their own in-house projects for storing, retrieving, organizing and managing annotated draft and complete genome sequence data. The software provides a web interface for the use of BLAST, allowing users to perform preliminary comparative analysis among multiple genomes. We demonstrate the utility of DraGnET for performing comparative genomics on closely related bacterial strains. Furthermore, DraGnET can be further developed to incorporate additional tools for more sophisticated analyses.

Conclusions: DraGnET is designed for use either by individual researchers or as a collaborative tool available through Internet (or Intranet) deployment. For genome projects that require genome sequencing data to initially remain proprietary, DraGnET provides the means for researchers to keep their data in-house for analysis using local programs or until it is made publicly available, at which point it may be uploaded to additional analysis software applications. The DraGnET home page is available at http://www.dragnet.cvm.iastate.edu and includes example files for examining the functionalities, a link for downloading the DraGnET setup package and a link to the DraGnET source code hosted with full documentation on SourceForge.

Figure 1

Java classes. Four Java classes are used to define the Gene, Strain, Logininformation and Blastdbupdate objects. The Gene class defines variables for the following annotated gene information: gene identification (gid and geneId), gene function (function), the protein sequence (proteinSequence), a description of the gene (geneDescription), the name of the gene (geneName), the size of the protein (proteinSize), the subcellular localization (localization), if the protein is predicted to be a lipoprotein (lipoprotein), if the protein is predicted to have a signal sequence (signalSequence) and the set of strains that contain the genes (Set<Strain> strains). The Strain class defines variables for strain information such as a strain identifier (sid and strainId), the strain name (strainName), a description of the strain (strainDescription), and the set of genes contained in the strain (Set<Gene> genes). The Logininformation class defines variables for the user login identifier (LogininformationId id), the usertype and the time the user logged in (lastlogon). The Blastdbupdate class defines variables for the date the last update was made to the data (dateId and updatedDate).

Figure 2

DraGnET software architecture. The DraGnET web application uses Struts to implement the Model-View-Controller (MVC) architecture. The view represents the presentation of the application and is implemented through Java Server Pages (JSP). The Controller is responsible for intercepting and translating user input into actions to be performed by the Model. The Controller receives the request from the browser, invokes a business operation and coordinates the view to be returned to the browser. The Struts Action servlet populates information from the JSP to the appropriate Struts Action Form then throws control to the Struts Action. The Struts Action gets data from the appropriate Struts Action Form and sends the information to the Model where certain actions like retrievals and updates will be performed. The Model is where communication with the database takes place through Hibernate. Hibernate is used to map Model Classes (Java objects) to tables in the database. Model Classes are also used to execute BLAST functionalities provided through the application's web interface. The Model represents enterprise data and the business rules that govern access to and updates of this data.

Figure 3

Web interface- DraGnET Home Page. Listed on the DraGnET home page are links for downloading the DraGnET setup package ("DraGnET Application Setup Package"), testing search and BLAST capabilities ("Example Files"), generating FASTA formatted files ("Generate FASTA files") and all "Search" functionalities.

Figure 4

Adding a new strain. The data entry tables displayed on the web pages for inserting a new strain. In the first table (A) the curator enters the strain Id, strain name and strain description of the new strain. In the second table (B) the curator is directed to upload a file containing gene information for genes contained in the strain. The strain and gene information is then stored in the database.

Figure 5

Data Modification. The table displayed on the web page for modifying gene and strain data. As shown in the table, modifications that can be made by the curator to gene and strain data include adding, deleting and updating gene or strain information.

Figure 6

Updating gene information. The data entry tables displayed on the web pages for updating gene information. In the top left table (A) the gene Id of the gene whose information needs modification is entered and submitted. The table on the bottom left (B) allows the curator to select gene attributes that need to be modified. Subsequently, as shown in table (C), for each attribute selected, the gene information currently stored in the database is displayed on the left side of the table as "Old" information and on the right side of the table changes to the gene information may be entered under "New".

Figure 7

Quick Search. The data entry tables and results table displayed on the web pages for "Quick Search". In the top left table (A) the user selects a gene or strain search attribute. In the bottom left table (B) the user enters information for the chosen search criteria. Subsequently, the results table (C) displays information for the chosen gene or strain.

Figure 8

Advanced Search. The data entry tables displayed on the web pages for "Advanced Search". Using the table on the left (A) users may customize their search for gene information by selecting single or multiple search attributes. The table on the right (B) allows users to enter and select values for the chosen search criteria. Subsequently, a text file containing search results is available for download.

Figure 9

BLAST Search. The data entry table and results displayed on the web pages for "BLAST Search". As shown in table (A), users have the option to select a single or multiple search database(s). In this example we have three BLAST databases available for searching that represent strains SH0165, 29755 and 12939. To refine their search, users have the option to change the E-value with the default being .01. A text box is provided for users to enter a FASTA formatted protein query sequence. The results (B) are displayed to the user and are available for download.

Figure 10

Batch BLAST Search. The data entry table displayed on the web page for "Batch BLAST Search". Users select a single search database from a drop down menu. In this example, the BLAST database representing strain 29755 was chosen. To refine their search, users have the option to change the E-value with the default being .01. Users then upload a text file containing FASTA formatted protein sequences which will be used as the set of query sequences. The results format is the same as "BLAST Search".