Lee H. Pratt

Bioinformatics
(click on any thumbnail for more detailed views)

MAGIC GENE DISCOVERY

The accompanying figure (click on this and other thumbnail views for a larger image) illustrates the primary components of that part of the database relating to sequence data.  While its focus is on ESTs, it also handles genomic sequencing and works with ad hoc sequencing projects such as high-throughput production of full-length cDNA sequences.

MAGIC ARRAY

The MAGIC ARRAY segment of the integrated database is MIAME compliant.  A user interacts with the database using MAGIC Array Data Manager, which is introduced below.  We maintain a Spotfire Server that permits exploration of microarray data over the web via a Spotfire browser (see below).  Oracle materialized views provide nearly instantaneous and always current annotation information about every spot in an array.

All activity in the wet lab is tracked via custom laboratory information management systems.  The front ends of both the sequencing pipeline (left) and the MAGIC Microarray Data Manager (right) are illustrated.  The latter is used remotely by collaborators who enter data, including TIF images, via the internet.  Spotfire (commercial visualization software) is used for data analyses, both locally and - via the internet - remotely.

Administration

An administrative interface permits a laboratory PI or supervisor to create user accounts and projects, assign view and data entry permissions, and define vectors, primers, genotypes, and so on.  All information required for high-throughput GenBank deposition is entered here.  This interface can be used remotely via the internet, permitting collaborators to utilize fully our local MAGIC Database.

Processing Pipeline

A novel modular, object-oriented pipeline for automated processing of sequence data coming from an ABI3700 or 3730XL has been created.  It can be used with cDNA sequences as well as shotgun (genomic) sequences.

Monitoring Wet-Lab Output

A laboratory PI or supervisor can easily track data production and the quality of that data via a pair of interfaces, one focusing on project status (left) and the other on sequence statistics (right).

Quality Control

Numerous visualization tools are provided to assist with quality control issues.  Illustrated here is one that monitors capillary performance of a sequencer by viewing read lengths as a function of user-selected limit dates.  This tool also tracks performance of individual capillaries as a function of time.

Sequence Viewer

Because we export the public contents of the MAGIC Database to this web site nightly, a number of our query and visualization tools can be used by everyone.  A Java Web Start program can be downloaded and used to view sequences.  Both text and graphical views of the data are available - the latter provides access to individual phred quality scores.  This tool also permits any viewer to download any or all sequences in fasta format, trimmed in a variety of user selectable ways.  Click on the image to the left to see more about this tool.  Click here to use this tool.

Gene Discovery

An even more powerful query/visualization tool that can be used at this web site is one we call Gene Discovery.  Following selection of a clustering run, it returns from the Oracle database to the user information about electronic expression profiles, contig alignments, sequence and contig annotation, and much more.  It also provides several query opportunities allowing any user, for example, to identify all contigs with at least 6 members, at least 4 of which have been detected only in libraries from drought-stressed sorghum.  Please click on the image to the right in order to learn more about this tool.  Please note that it is gaining function over time, such that what you see on the web site might appear somewhat different, and almost certainly will have functions beyond those mentioned here.  Click here to use this tool.