Computational Model of Gene Regulatory Program

点击翻译按钮获取中文摘要

PROJECT SUMMARY / ABSTRACT Gene regulatory networks (GRN) are models which represent the genomic regulatory code that directly controls the processes of embryonic development, and also the postembryonic construction of the body plan, including organogenesis. The most advanced of these models at present is the sea urchin endomesoderm GRN. This GRN specifies the interactions of regulatory genes that are the direct output of the regulatory code, and that in terms of regulatory logic lie upstream of all other aspects of the regulatory system including microRNAs, chromatin modifications and transcription biochemistry. In the current period of the Grant for which we are seeking renewal, we developed BioTapestry as the computational platform for representation of the endomesoderm GRN, including direct functional significance of regulatory DNA sequence, architecture of the genomically encoded control system, temporal and spatial activity of GRN interactions, and extensive linkage to the underlying experimental data. BioTapestry is now the leading vehicle for GRN analysis and presentation. The computational aspect of the project supported by this grant has been inseparably intertwined with experimental exploration of GRN properties, and the same is true of this renewal proposal. New computational objectives include a vast enhancement of BioTapestry so that it will have the unique capacities of computing GRN architecture from experimental data, and indicating to the investigator aspects in need of further information. A second computational objective is development of a set of methods for utilizing time course measurements of activities of genes within the GRN to compute the kinetic behavior of the subcircuits of the GRN that execute particular developmental functions, and testing the results vs. observed data. We will also develop computational apparatus for assessing the sufficiency of the GRN in explaining the observed patterns of gene expression. However, sufficiency must ultimately be assessed experimentally. Two major initiatives are proposed to interlock with the computational assessment of sufficiency of explanation: first, we will develop entirely new multiplexed methods for cis-regulatory module identification and experimental analysis, and we will use these to verify predictions of the GRN at all or most GRN nodes. As an ultimate challenge to the sufficiency of the explanatory power of the GRN, we will use a synthetic approach, by constructing regulatory subcircuits in re-engineered BACs, that should suffice to produce a predicted developmental outcome. They will be inserted into sea urchin eggs and sea star eggs to determine whether they indeed generate the predicted developmental patterns of gene expression in contexts where they would not normally appear. PROJECT NARRATIVE This is a basic research project directed at the fundamental challenge of understanding what animal genomes mean, in particular that aspect of life process by which animals develop according to evolutionarily inherited, species specific genomic programs. Genomic control systems are complex logic processing mechanisms that require custom designed, computational apparatus for understanding and analysis. We will create new potentialities in what is already the world's leading computational platform for handling these control systems, and we will experimentally test the sufficiency of our understanding of them, a project that will be immediately relevant to all animal genomic control systems including our own.