RESEARCH:

Bacterial viruses (phage) are the most abundant and diverse biological entities on Earth (phage abundance/diversity is ~10x greater than that of prokaryotes, which in turn are ~100x greater than that of eukaryotes). Phage preform a vital role in environmental nutrient cycles through how they interact with their bacterial hosts. These interactions can take one of a few different routes, both ultimately culminating in the replication of phage. During bacterial infection, phage modulate their host’s metabolism and energy dynamics, which can result in changing host population size by either causing outright death of the host or providing a physiological advantage promoting host survival.

Even with an understanding of the importance, abundance and diversity of phage and their effect on nutrient cycling, we know relatively little about phage gene content; roughly 70% of all phage genes are of unknown function. Of the genes that have been characterized, they can be classified into one of three different functional groups: (i) phage replication, (ii) phage structural proteins, (iii) proteins that affect host metabolism. Genes involved in phage replication tend to be well conserved and are relatively easy to identify based on sequence similarity searches, while phage structural and metabolic genes tend to be more variable. Computational methods (e.g., artificial neural networks) are being employed to identify potential structural proteins, to much success, while metabolic proteins require more wet lab-based identification methodologies. Through working with Drs. Rob Edwards, Forest Rohwer and Anca Segall, I have developed a pipeline to predict functions for phage metabolic genes through the combination of physiological methods (phenotype microarrays and metabolomics) and genomics (metabolic modeling and metabolic networks) by creating a computational interface between biological and genomic information. Functional predictions are subsequently verified through the use of bacterial genetics and physiology.

Bacterial viruses (phage) are the most abundant and diverse biological entities on Earth (phage abundance/diversity is ~10x greater than that of prokaryotes, which in turn are ~100x greater than that of eukaryotes). Phage preform a vital role in environmental nutrient cycles through how they interact with their bacterial hosts. These interactions can take one of a few different routes, both ultimately culminating in the replication of phage. During bacterial infection, phage modulate their host’s metabolism and energy dynamics, which can result in changing host population size by either causing outright death of the host or providing a physiological advantage promoting host survival.

Even with an understanding of the importance, abundance and diversity of phage and their effect on nutrient cycling, we know relatively little about phage gene content; roughly 70% of all phage genes are of unknown function. Of the genes that have been characterized, they can be classified into one of three different functional groups: (i) phage replication, (ii) phage structural proteins, (iii) proteins that affect host metabolism. Genes involved in phage replication tend to be well conserved and are relatively easy to identify based on sequence similarity searches, while phage structural and metabolic genes tend to be more variable. Computational methods (e.g., artificial neural networks) are being employed to identify potential structural proteins, to much success, while metabolic proteins require more wet lab-based identification methodologies. Through working with Drs. Rob Edwards, Forest Rohwer and Anca Segall, I have developed a pipeline to predict functions for phage metabolic genes through the combination of physiological methods (phenotype microarrays and metabolomics) and genomics (metabolic modeling and metabolic networks) by creating a computational interface between biological and genomic information. Functional predictions are subsequently verified through the use of bacterial genetics and physiology.