GRiTS

GRITS- Gene Regulation in Time and Space.

The goal of the GRITS project is to study gene expression information in three dimensions and over a period of time. We have developed a suite of software tools that will allows researchers to correlate gene information with respect to both time and space. Currently there are two applications in development. The first,GRITS Imagetool , lets the user segment the gene image data by density as a function of spatial location. The data is stored relative to its “stage” (the time coordinate) , user supplied data, and its position within the volume(x,y,z).

The GRITS Viewer gives the user the ability to select, view and search the available data sets. Queries about gene presence or correlations with other genes can be performed in several ways. One is by co-expression within a defined volume. The user selects a point within the data set and then defines a volume around that point that defines the search. A search is then performed across all of the gene expression data stored in the published database and returns those genes that lie within the defined volume. The second method for searching the data is done using meta data provided by those who in-put the data. The third method is to perform a combined query using equivalent space(volume) information between stages with user supplied information. Information such as stage, image date, institution, gene type can be dynamically added to a query. For example,data can be searched within the equivalent volume and for a set of stages (P01,P02,P03). A future enhancement to the viewer tool will animate the results of the search, displaying the genes found in the volume over the various stages.

How it works.

Each of the applications is a standalone application residing on the client’s desktop. A central server handles data storage and retrieval. The GRITS viewer can either cache data retrieved from sever (good for giving a demo where there internet connection), or allow for on-line use. The GRITS Imagetool processes image files locally and uploads them to the central database, once the user is satisfied with the accuracy of the data. The central server utilizes the JBoss application server, Hibernate persistence framework, and MySQL database. Client server communication is done via secure SOAP(Web Services).

The applications.

The client applications are built using the Visualization Tool Kit(VTK) with QT. The later was selected for it cross platform capabilities, though this version of the applications run only on the Windows platform.

GRITS Imagetool.

Below is a view of the imageTool and one processed gene image. The red represents the sample gene information and the blue represents the boundary of the original image. The tissue boundary is used register different data sets to a standardized model representing both developmental stage and relative spatial coordinates.

Before the data can be uploaded to the server one must first log in. Until then the “Upload Gene” button will be disabled. No actions related to the database can be performed without logging in.

The image tool allows the user to define up to five density ranges. For each range the user selects start and end values along with a color value. The “sphere” setting indicates the size of the sphere to be displayed.

imagetoolfilter

Each set of data can be associated with additional information, “metadata”. The Imagetool application allows the user to define this data but also add their own metadata keywords. The GRITS Viewer application will display a list of the defined keywords that can be used in the query.

keywords

Keywords can be added, modified or deleted. The “Delete” button shown here only removes a row from the local table. To protect the integrity of the data, deletion of keyword from the database can only be done by the administrator .

The user can assign values specific key words

metadata

The viewer

gritsviewer1

gritsviewerfloat

gritsviewerfloat2

Model slicing

The blue region represents a single “slice” of the standardized model

model slicing

Model and data scaling

The classified data(red,green) has been scaled to fit the dimensions of the standardized model for that specific developmental stage.

image scaling 1

The spatially adjusted data is stored in a central database for future viewing, along with the original image data.