Developmental Biology and Biological Modeling

    The long-term objective is to integrate the masses of genetic and molecular data that have been generated over years of reductionist experimentation into a coherent understanding of the developmental process. With the birth of the molecular biology revolution and the subsequent exponential increase in molecular genetic data, it has become possible for biologists to incorporate only smaller and smaller portions of this information into their hypotheses; gains in understanding are not keeping pace with gains in knowledge. It is largely the success of the reductionist approach in revealing the complexity of the system that has made apparent the need to develop and employ computational tools to enable the understanding of how the system functions as a whole.

One of the primary driving forces behind the explosion of data is the genome projects that were the inevitable consequence of the molecular biology revolution. Genome projects are providing a huge excess of easy "leads" to explore, and a new bottleneck in the scientific pipeline has arisen centering around data analysis.

The aims of this project are:

1) To create an agent-based computational modeling and simulation system for multicellular tissues. This software suite will enable the modeling of biochemical and cellular processes with a level of detail and realism previously not possible. Individual cells, organelles, genes, and gene products will be represented in a biologically meaningful fashion, enabling one to visualize the effects of various perturbations (e.g. mutations, drugs, etc.) within the context of a complex signaling network over the course of development.

2) To use this system to model the embryonic development of the Drosophila melanogaster wing. The entire Drosophila developmental regulatory system will be modeled, with every gene known or suspected to be involved in patterning and signaling explicitly represented (estimated to be 200 - 800 genes), incorporating diverse types of data into a maximally self-consistent model. The developmental process, beginning with the early imaginal disc primordia (~50 cells) will be modeled through mid-third instar, when the wing blade is fully patterned, but has yet to differentiate (~2,000-10,000 cells).

3) To deduce previously unknown regulatory interactions of individual genes or signaling pathways necessary for models to result in gene expression patterns and morphology consistent with wild-type. This modeling system will identify likely points of cross-talk among the several signaling pathways, along with currently unexpected roles for products of known genes.

4) To test the validity of the model and utility of the modeling system by examining those proposed regulatory interaction in the laboratory. The phenotypes of novel mutation combinations will be modeled in silico, and interesting or unexpected results will be reproduced experimentally using available mutant strains as a rigorous test of model fidelity. In addition, theoretical gene expression profiles at various points in development will be tested for many genes simultaneously through the use of expression microarrays.