Intramural Research
Online Research Resources
WebBLAST 1
|
WebBLAST: An Integrated Solution for Organizing and Analyzing Sequence Data
This is the Abstract for WebBLAST1.
Joseph Ryan, Erik Ferlanti, Izabela Makalowska, and Andreas D. Baxevanis. Genome Technology Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, Maryland 20892 USA.
WebBLAST is a suite of Macintosh and UNIX programs intended to assist in organizing sequencing data as well as providing first-pass sequence analysis in an automated fashion. One of the main features of WebBLAST is that its system requirements are minimal, needing only MacPerl on the Macintosh client, Perl 5.004 and BLAST on the UNIX client, and a Web server. All output is viewed through a Web browser, allowing users to access the refined data regardless of platform.
Data enters the WebBLAST pipeline through the use of a MacPerl droplet resident on the same Macintosh to which the ABI sequencer is connected. The droplet transfers the data into the UNIX file system and places the data into a holding area on the server. A nightly cron job then invokes the automatic processing of new data: [1] Sequences are moved from the holding area to the appropriate directory within the UNIX file system. [2] Sequences are screened for the presence of vector, alu, and other repetitive sequences and masked. [3] Local BLAST searches are then run on the masked data, and individual BLAST hits are assessed for statistical significance. [4] Based on user-configurable prerequisites, the BLAST statistics are evaluated by a CGI program, which then generates HTML pages presenting the user with a summary table of hit counts, as well as information on the significant hits. [5] Sequence data are re-BLASTed monthly against the NCBI month database, with any new hit information being appended to the summary pages of the appropriate sequence set; this assures that BLAST results are kept up-to-date.
Once the sequence comparisons have been run, users have the ability to add comments to individual sequences that can be viewed by all users who have priveleges to view that data set. Users can also view the associated sequence traces directly through the use of a trace viewer, a Java applet which can be invoked automatically by clicking on a hyperlink. Each data set can be password-protected, so that multiple projects can be supported while maintaining the integrity and security of the data. An export function is available, so that data can be downloaded for use with assembly programs such as the PHRED/PHRAP/ CONSED suite. As the programs are written in perl, users can easily modify the scripts to include additional analyses, such as performing BLAST runs against in-house data sets.
11th Annual Cold Spring Harbor Meeting on Genome Mapping, Sequencing, and Biology, May 1998
Comments, suggestions and problems to bioinformatics@nhgri.nih.gov
