| |
John Reinitz, Ph.D.
Associate Professor of Applied Math and Statistics;
Investigator Center for Developmental Genetics.
Funding through the National Center for Research Resources.
Figure 1. Results of immunostaining, image processing,
and data acquisition. A) Drosophila
embryo triple stained for Giant (Gt; red), Kruppel (Kr; green),
and Even-skipped (Eve; blue) protein. Each colored dot is a
blastoderm nucleus. Nuclei which contain more than one of these
proteins appear as blends of two colors. B)
Binary image of the nuclear mask. Manipulation of the image
in A results in the definition of image domains which correspond
to individual nuclei (white dots). Each domain is bounded by
pixels in the "off" state (black background). C)
Graph of expression data for Gt, Kr, and Eve from the central
strip of the embryo image. Horizontal lines (white in A, red
in B) demarcate the zone of nuclei from which intensity data
was calculated. Average pixel intensity per nucleus is plotted
against anterior-posterior position. Gt: dotted line, Kr: dashed
line, and Eve: solid line. |
|
My laboratory is engaged in a long term project to characterize the physiology
of the segment determination process in the fruit fly Drosophila melanogaster
by means of experiment, computation, and mathematics. The major goal of
our current work is to achieve an integrative understanding of the segment
determination process as an emergent property of a network of genes, a
copy of which is present in each of many cells. This research will provide
a precise description of determination and regulation in a morphogenetic
field in terms of physiology and chemistry. This goal was first envisaged
over a hundred years ago, and is finally achievable now.
To accomplish this goal requires the integrated application of a set
of techniques from many disciplines on a single biological problem. We
obtain spatially localized gene expression data in the form of confocal
images and process it through a series of steps into a numerical atlas
of gene expression at cellular resolution. This data is currently available
on the web at http://urchin.spbcas.ru/flyex;
the proper design and construction of this database is itself a frontier
problem in bioinformatics. The gene expression data is then fit to a set
of nonlinear ODE's by large scale numerical optimization on parallel computers.
Finally, we try to understand new biology from the results of these fits.
We are exploring a version of the model expressed as PDE's in an effort
to gain insight into the problem by analytical mathematics.
All of the activities mentioned above involve interesting research questions.
An engineering or physical science student's project in the laboratory
could involve any of the areas listed, depending on the student's skills.
Because of the diverse methodologies we use, the project really depends
on the student. Possible projects range from improvement of numerical
methods of integration to database replication to the testing of new flurophores.
The only essential background requirement is interest and the willingness
to learn new things. Interested students should contact the PI for further
discussion.
Contact Information
email: reinitz@ams.sunysb.edu
url: http://flyex.ams.sunysb.edu
back to top
|
|
