Deok-Sun Lee

The implications of human metabolic network topology for disease comorbidity

Most diseases are the consequence of the breakdown of cellular processes, but the relationships among genetic/epigenetic defects, the molecular interaction networks underlying them, and the disease phenotypes remain poorly understood. To gain insights into such relationships, here we constructed a bipartite human disease association network in which nodes are diseases and two diseases are linked if mutated enzymes associated with them catalyze adjacent metabolic reactions. We find that connected disease pairs display higher correlated reaction flux rate, corresponding enzyme-encoding gene coexpression, and higher comorbidity than those that have no metabolic link between them. Furthermore, the more connected a disease is to other diseases, the higher is its prevalence and associated mortality rate. The network topology-based approach also helps to uncover potential mechanisms that contribute to their shared pathophysiology. Thus, the structure and modeled function of the human metabolic network can provide insights into disease comorbidity, with potentially important consequences for disease diagnosis and prevention.

The Impact of Cellular Networks on Disease Comorbidity

The impact of disease‐causing defects is often not limited to the products of a mutated gene but, thanks to interactions between the molecular components, may also affect other cellular functions, resulting in potential comorbidity effects. By combining information on cellular interactions, disease‐‐gene associations, and population‐level disease patterns extracted from Medicare data, we find statistically significant correlations between the underlying structure of cellular networks and disease comorbidity patterns in the human population. Our results indicate that such a combination of population‐level data and cellular network information could help build novel hypotheses about disease mechanisms.

Comparative Genome-Scale Metabolic Reconstruction and Flux Balance Analysis of Multiple Staphylococcus aureus Genomes Identify Novel Antimicrobial Drug Targets

Mortality due to multidrug-resistant Staphylococcus aureus infection is predicted to surpass that of human immunodeficiency virus/AIDS in the United States. Despite the various treatment options for S. aureus infections, it remains a major hospital- and community-acquired opportunistic pathogen. With the emergence of multidrug-resistant S. aureus strains, there is an urgent need for the discovery of new antimicrobial drug targets in the organism. To this end, we reconstructed the metabolic networks of multidrug-resistant S. aureus strains using genome annotation, functional-pathway analysis, and comparative genomic approaches, followed by flux balance analysis-based in silico single and double gene deletion experiments. We identified 70 single enzymes and 54 pairs of enzymes whose corresponding metabolic reactions are predicted to be unconditionally essential for growth. Of these, 44 single enzymes and 10 enzyme pairs proved to be common to all 13 S. aureus strains, including many that had not been previously identified as being essential for growth by gene deletion experiments in S. aureus. We thus conclude that metabolic reconstruction and in silico analyses of multiple strains of the same bacterial species provide a novel approach for potential antibiotic target identification.

Blueprint for antimicrobial hit discovery targeting metabolic networks

Advances in genome analysis, network biology, and computational chemistry have the potential to revolutionize drug discovery by combining system-level identification of drug targets with the atomistic modeling of small molecules capable of modulating their activity. To demonstrate the effectiveness of such a discovery pipeline, we deduced common antibiotic targets in Escherichia coli and Staphylococcus aureus by identifying shared tissue-specific or uniformly essential metabolic reactions in their metabolic networks. We then predicted through virtual screening dozens of potential inhibitors for several enzymes of these reactions and showed experimentally that a subset of these inhibited both enzyme activities in vitro and bacterial cell viability. This blueprint is applicable for any sequenced organism with high-quality metabolic reconstruction and suggests a general strategy for strain-specific antiinfective therapy.

Flow correlated percolation during vascular remodeling in growing tumors.

A theoretical model based on the molecular interactions between a growing tumor and a dynamically evolving blood vessel network describes the transformation of the regular vasculature in normal tissues into a highly inhomogeneous tumor specific capillary network. The emerging morphology, characterized by the compartmentalization of the tumor into several regions differing in vessel density, diameter, and necrosis, is in accordance with experimental data for human melanoma. Vessel collapse due to a combination of severely reduced blood flow and solid stress exerted by the tumor leads to a correlated percolation process that is driven towards criticality by the mechanism of hydrodynamic vessel stabilization.

Synchronization transition in scale-free networks: clusters of synchrony.

We study the synchronization transition in scale-free networks that display power-law asymptotic behaviors in their degree distributions. The critical coupling strength and the order-parameter critical exponent derived by the mean-field approach depend on the degree exponent lambda, which implies a close connection between structural organization and the emergence of dynamical order in complex systems. We also derive the finite-size scaling behavior of the order parameter, finding that the giant cluster of synchronized nodes is formed in different ways between scale-free networks with 2 < lambda < 3 and those with lambda > 3.

First passage time for random walks in heterogeneous networks.

The first passage time (FPT) for random walks is a key indicator of how fast information diffuses in a given system. Despite the role of FPT as a fundamental feature in transport phenomena, its behavior, particularly in heterogeneous networks, is not yet fully understood. Here, we study, both analytically and numerically, the scaling behavior of the FPT distribution to a given target node, averaged over all starting nodes. We find that random walks arrive quickly at a local hub, and therefore, the FPT distribution shows a crossover with respect to time from fast decay behavior (induced from the attractive effect to the hub) to slow decay behavior (caused by the exploring of the entire system). Moreover, the mean FPT is independent of the degree of the target node in the case of compact exploration. These theoretical results justify the necessity of using a random jump protocol (empirically used in search engines) and provide guidelines for designing an effective network to make information quickly accessible.

A Mapping of Drug Space from the Viewpoint of Small Molecule Metabolism

Small molecule drugs target many core metabolic enzymes in humans and pathogens, often mimicking endogenous ligands. The effects may be therapeutic or toxic, but are frequently unexpected. A large-scale mapping of the intersection between drugs and metabolism is needed to better guide drug discovery. To map the intersection between drugs and metabolism, we have grouped drugs and metabolites by their associated targets and enzymes using ligand-based set signatures created to quantify their degree of similarity in chemical space. The results reveal the chemical space that has been explored for metabolic targets, where successful drugs have been found, and what novel territory remains. To aid other researchers in their drug discovery efforts, we have created an online resource of interactive maps linking drugs to metabolism. These maps predict the “effect space” comprising likely target enzymes for each of the 246 MDDR drug classes in humans. The online resource also provides species-specific interactive drug-metabolism maps for each of the 385 model organisms and pathogens in the BioCyc database collection. Chemical similarity links between drugs and metabolites predict potential toxicity, suggest routes of metabolism, and reveal drug polypharmacology. The metabolic maps enable interactive navigation of the vast biological data on potential metabolic drug targets and the drug chemistry currently available to prosecute those targets. Thus, this work provides a large-scale approach to ligand-based prediction of drug action in small molecule metabolism.

Synchronization transition of heterogeneously coupled oscillators on scale-free networks.

We investigate the synchronization transition of the modified Kuramoto model where the oscillators form a scale-free network with degree exponent lambda . An oscillator of degree k_{i} is coupled to its neighboring oscillators with asymmetric and degree-dependent coupling in the form of Jk_{i};{eta-1} . By invoking the mean-field approach, we find eight different synchronization transition behaviors depending on the values of eta and lambda , and derive the critical exponents associated with the order parameter and the finite-size scaling in each case. The synchronization transition point J_{c} is determined as being zero (finite) when eta>lambda-2 (eta<lambda-2) . The synchronization transition is also studied from the perspective of cluster formation of synchronized vertices. The cluster-size distribution and the largest cluster size as a function of the system size are derived for each case using the generating function technique. Our analytic results are confirmed by numerical simulations.

Distribution of extremes in the fluctuations of two-dimensional equilibrium interfaces.

We investigate the statistics of the maximal fluctuation of two-dimensional Gaussian interfaces. Its relation to the entropic repulsion between rigid walls and a confined interface is used to derive the average maximal [EQUATION: SEE TEXT]and the asymptotic behavior of the whole distribution for [EQUATION: SEE TEXT] for m finite with N2 and K the interface size and tension, respectively. The standardized form of P(m) does not depend on N or K, but shows a good agreement with Gumbels first asymptote distribution with a particular noninteger parameter. The effects of the correlations among individual fluctuations on the extreme value statistics are discussed in our findings.