This project aims to 1) characterize large-scale protein-protein interaction (PPI) networks of specific organisms, 2) compare large-scale PPI networks of related organisms, 3) understand the fundamental properties of protein and protein-complex interactions, and to 4) understand how host-pathogen PPIs leads to host infection. This knowledge will aid our understanding of how biological processes are organized, lend insights into the mechanisms of various biological actions, and identify proteins and interactions that are responsible for disease.

Characterize large-scale PPI networks

This objective aims to discover the inherent natures of PPI networks of organisms, including those that are pathogenic. This research will provide insight into the architectural structure of the PPI networks and elucidate the ways in which the large numbers (in the thousands) of constituent proteins interact amongst themselves. Such information will help bridge the gaps in understanding how proteins combine to result in overall organism behavior, such as lethality. One of the major impacts of this research will be to identify how probabilistic in nature (or how random) biological networks are and conversely, discover whether (and if so where) biological guiding principles are evident.

Our research has elucidated very interesting and common characteristics of these PPI networks, indicating that their wiring appears to be largely random in nature. Whether this apparent randomness is the result of the evolution process or some other underlying phenomenon or experimental bias is currently under investigation.

Compare large-scale protein-protein interaction networks

This objective is focused on explaining the similarities and differences in behavior between related organisms. For example, Escherichia coli and Yersinia pestis (plague) are considered to be evolutionary neighbors. However, they have quite different virulence. This objective will be realized by comparing biological PPI networks of related pathogens. Through these comparisons we aim to identify specific regions that are conserved and conversely, those regions that are distinct. Further analysis of these conserved and distinct regions will provide insight into the analogous and contrastive modes of action of the pathogens.

Fundamental properties of protein and protein-complex interaction

It is well known that PPIs determined from different experiments, for the same species, show a small number of overlapping interactions. The reasons for this are not clear. As proteins contain domains, or regions with well-characterized functional/structural properties, it may be more sensible to study PPI networks through their underlying domain-domain interaction (DDI) networks. Work in progress has shown that DDIs from different experiments are more consistent than their corresponding PPIs. Therefore, DDI networks may contain more relevant information, leading to enhanced understanding of the proteins' network attributes. For example, presence of a particular domain may lead to a protein having very many interactions.

Host-pathogen PPIs

Possession of accurate DDI profiles for one organism can lead to enhanced inference, or prediction, of PPIs for another species. As the experimental determination of a complete PPI network for a given organism requires vast resources, accurate methodologies would be of great value. One use of this prediction technology would be to investigate the interactions that occur between a pathogen, such as Y. pestis, with human host cells in terms of their PPIs and gain insights into the mechanisms of host invasion and subsequent impairment.