The Bioinformatic Software Bench (BSB) for Drupal allows for a rapid and secure deployment of bioinformatic software. The BLAST, InterProScan, SSAHA2 and annot8r software are provided by default and plugins for new software can be readily written. An innovative graphical user interface guides both end-user and administration of the software, including the secure provision of pre-publication datasets.
BSB is written for Drupal 6 and it depends on the Drupal gmod_dbsf module (http://drupal.org/project/gmod_dbsf). The NCBI BLASTALL package and other relevant software for each plugin are required in order to provide the relevant facilities but administrators can choose which software to enable.
It has a number of attractive features:
For the system administrator:
- Secure & fast to deploy: the demonstration required 30' to setup from scratch, 10' of which was the BLAST server. The administrator has to merely click on a checkbox to activate the software and then choose which databases will be available to the plugins.
- Secure with private data: We know that laboratories often have two types of data: those they wish to make public and the unpublished ones they wish to utilize only in-house. The solution is often the creation of two separate servers, one password protected. The DBFS takes care of user management and allows you to choose which datasets are available to which users. An unlimited number of user groups can be created included the anonymous/guest visitor who has not logged in.
- Plugin format: It comes with NCBI's BLASTALL but DBFS is an API framework and easy to extend, our publication will have a manual as supplementary material to guide bioinformaticians in writing their plugins with 1-2 days.
- High Performance Computing (HPC) support: Optionally, it can utilise the Condor job management system so jobs run on a PC-farm instead of the often meek BLAST server. A perl daemon manages the submissions automatically.
- Database API is Chado-compatible: It is GMOD compatible and uses the robust, normalized Chado database. As Chado is sequence feature-driven we created sequence-less table set which follows Chado-convention. The future aim is to be able to integrate multiple website deploying this software.
For the end-user:
- Build for wet-lab researchers. They can upload a multi-FASTA file to be used as the query or even the reference database: there is no need to restrict the user to the administrator's datasets. The file size limit is customizable. Users can bookmark their search and re-retrieve it within 7 days. Further, they can simultaneously run multiple BLAST algorithm searches.
- Driven by open-source software and adheres to community standards: The output is driven by BioPerl? (XML, HTML, TXT and a graph using Bio::Graphics). The database infrastructure is GMOD-compatible and the visualization is securely managed by Drupal.
Demos and Screenshots
This software is GPLv2+ licensed as part of the Drupal package. It is currently supported by Alexie Papanicolaou while at CSIRO, the Australian Commonwealth Scientific and Research Organization (http://www.csiro.au).