Networks of interacting proteins mediate a diverse range of biological processes. An essential step toward developing a more thorough understanding of biological systems is to catalogue all of the binary interactions that comprise these protein networks for each genome. The long-term objective of this project is to construct a comprehensive high-quality protein interaction map for the genetically tractable model organism, Drosophila melanogaster. A proven technology for detecting biologically informative protein interactions is the yeast two-hybrid system. In a completed phase of this project, yeast two-hybrid clone arrays were constructed for over 86% of the 14,000 predicted Drosophila genes. Initial screening of these arrays detected thousands of protein interactions, which are proving invaluable for understanding protein and pathway function. Analyses have also revealed that only a small fraction of the interactions detectable with this technology have been identified thus far.
The first aim of this project is to increase the coverage of protein interaction data by continued two-hybrid screening, particularly with the proteins missing from current interaction maps;this will at least double the existing number of interactions. In the second aim, two different assay systems will be used to verify all detected interactions. Using these results and other criteria, each interaction will be assigned a confidence score indicating the probability that it plays a functional role in vivo. In the third aim, co-affinity purifications will be performed to further validate the interaction data.
In Aim 4, binary interactions will be directly tested for groups of proteins suspected of being functionally related or between proteins computationally predicted to interact. Compared to random screening, this approach is likely to produce more complete and accurate interaction data for these proteins, and moreover, should provide data to help validate and further enhance prediction algorithms. In the final aim, newly predicted genes missing from the existing two-hybrid arrays will be cloned and incorporated into the screening process. All data from this project will be publicly available in standardized formats via at least two independent web-based databases. Combined, the data from this project will provide a foundation for understanding the functions of conserved protein networks and for modeling complex biological processes and diseases.
Showing the most recent 10 out of 13 publications
