Welcome to Computational Systems Biology and Bioinformatics Lab!

 

Home

Publications

People

Resources

Teaching

Directions

Dennis Vitkup, Ph.D.

Associate Professor
Center for Computational Biology and Bioinformatics
Department of Biomedical Informatics
1130 St. Nicolas Avenue
Columbia University
New York, NY 10032

Phone: 212-851-5151
Fax: 212-851-5149
E-mail: vitkup AT dbmi [dot] columbia [dot] edu

Research

Our lab is primarily interested in three main challenges of systems biology: reconstructing biological networks, understanding the evolution of biological networks, and developing efficient methods to simulate networks. The goal is to develop state-of-the-art computational tools and make specific predictions to be tested by our experimental collaborators.

Reconstruction of biological networks is an essential first step in systems biology. Currently, partial information on protein-protein and metabolic interaction networks is available for several model organisms (such as E. coli, S. cerevisiae, D. melanogaster, C. elegans). This information has been accumulated over several decades mainly based on careful experimental analysis. A major challenge is to accurately extrapolate the information from known networks to networks of other medically and biologically important organisms. We are currently pursuing these reconstruction efforts using probabilistic data integration and comparative genomics. To this end, we have developed a novel method which effectively combines the local structure of a metabolic network with context-based (co-expression, chromosomal distance, co-evolution) and sequence homology information to predict genes responsible for orphan metabolic activities. Currently about 30-40% of metabolic activities known to exist in sequenced organisms remain orphan, i.e. no sequences are assigned to these activities. Our computational method is able to successfully annotate about 60% of these activities. The work to extend our approach to signaling networks is in progress.

(click to enlarge picture)

(click to enlarge picture)

Evolution of biological networks and the evolution of proteins in the context of biological networks. Two specific questions are currently under investigation. First, we are interested in how biological modules and pathways evolve to perform new functions. Second, we are interested in establishing how the structure and properties of biological networks shape the evolution of network components. Our results demonstrate that, in yeast, important evolutionary processes, such as single nucleotide mutations, gene duplications, and gene deletions are influenced by the structure and function of the metabolic network. Specifically, central and highly connected enzymes evolve more slowly than less connected enzymes. Also, enzymes carrying high metabolic fluxes under natural biological conditions experience stronger evolutionary constraints.Duplications of genes encoding enzymes with high connectivity and high metabolic flux have a greater chance of being retained in evolution. In contrast to the case in protein interaction networks, highly connected enzymes are no more likely to be essential compared to less connected enzymes. These initial results demonstrate that systems biology approaches will be essential for understanding molecular evolution of enzymes.

 

Computer simulations of biological networks complement biological experiments. At present, the topological structure of biological networks is known much better than their kinetic and equilibrium parameters. Consequently, conceptual simplifications are required to obtain meaningful results from simulations. We are currently pursuing such studies in several specific systems. We have shown that two general principles govern the network behavior of bacterial metabolic networks: the optimality of the flux distribution in the native state and the dominance of homeostatic regulation upon perturbation. We demonstrated how these properties can be used to predict in-silico (using Linear and Quadratic optimization) the effects of gene deletions in E. coli and S. cerevisiae. Our results have been confirmed by several experimental laboratories. We are currently extending these studies to other organisms and perturbation types.

(click to enlarge picture)

Should you have any comments on the website, send an email to the webmaster.