Allison P. Heath

The temporal response of the Mycobacterium tuberculosis gene regulatory network during growth arrest

The virulence of Mycobacterium tuberculosis depends on the ability of the bacilli to switch between replicative (growth) and non‐replicative (dormancy) states in response to host immunity. However, the gene regulatory events associated with transition to dormancy are largely unknown. To address this question, we have assembled the largest M. tuberculosis transcriptional‐regulatory network to date, and characterized the temporal response of this network during adaptation to stationary phase and hypoxia, using published microarray data. Distinct sets of transcriptional subnetworks (origons) were responsive at various stages of adaptation, showing a gradual progression of network response under both conditions. Most of the responsive origons were in common between the two conditions and may help define a general transcriptional signature of M. tuberculosis growth arrest. These results open the door for a systems‐level understanding of transition to non‐replicative persistence, a phenotypic state that prevents sterilization of infection by the host immune response and promotes the establishment of latent M. tuberculosis infection, a condition found in two billion people worldwide.

From coarse-grain to all-atom: Toward multiscale analysis of protein landscapes

Multiscale methods are becoming increasingly promising as a way to characterize the dynamics of large protein systems on biologically relevant time‐scales. The underlying assumption in multiscale simulations is that it is possible to move reliably between different resolutions. We present a method that efficiently generates realistic all‐atom protein structures starting from the Cα atom positions, as obtained for instance from extensive coarse‐grain simulations. The method, a reconstruction algorithm for coarse‐grain structures (RACOGS), is validated by reconstructing ensembles of coarse‐grain structures obtained during folding simulations of the proteins src‐SH3 and S6. The results show that RACOGS consistently produces low energy, all‐atom structures. A comparison of the free energy landscapes calculated using the coarse‐grain structures versus the all‐atom structures shows good correspondence and little distortion in the protein folding landscape. Proteins 2007.

Survey: Computational challenges in systems biology

Systems biology is a broad field that incorporates both computational and experimental approaches to provide a system level understanding of biological function. Initial forays into computational systems biology have focused on a variety of biological networks such as protein-protein interaction, signaling, transcription and metabolic networks. In this review we will provide an overview of available data relevant to systems biology, properties of biological networks, algorithms to compare and align networks and simulation and modeling techniques. Looking towards the future, we will discuss work on integrating additional functional information with biological networks, such as three dimensional structures and the complex environment of the cell. Combining and understanding this information requires development of novel algorithms and data integration techniques and solving these difficult computational problems will advance both computational and biological research.

Finding metabolic pathways using atom tracking

MOTIVATION Finding novel or non-standard metabolic pathways, possibly spanning multiple species, has important applications in fields such as metabolic engineering, metabolic network analysis and metabolic network reconstruction. Traditionally, this has been a manual process, but the large volume of metabolic data now available has created a need for computational tools to automatically identify biologically relevant pathways. RESULTS We present new algorithms for finding metabolic pathways, given a desired start and target compound, that conserve a given number of atoms by tracking the movement of atoms through metabolic networks containing thousands of compounds and reactions. First, we describe an algorithm that identifies linear pathways. We then present a new algorithm for finding branched metabolic pathways. Comparisons to known metabolic pathways demonstrate that atom tracking enables our algorithms to avoid many unrealistic connections, often found in previous approaches, and return biologically meaningful pathways. Our results also demonstrate the potential of the algorithms to find novel or non-standard pathways that may span multiple organisms. AVAILABILITY The software is freely available for academic use at: http://www.kavrakilab.org/atommetanet. SUPPLEMENTARY INFORMATION Supplementary data are available at Bioinformatics online.

Defection detection: predicting search engine switching

Searchers have a choice about which Web search engine they use when looking for information online. If they are unsuccessful on one engine, users may switch to a different engine to continue their search. By predicting when switches are likely to occur, the search experience can be modified to retain searchers or ensure a quality experience for incoming searchers. In this poster, we present research on a technique for predicting search engine switches. Our findings show that prediction is possible at a reasonable level of accuracy, particularly when personalization or user grouping is employed. These findings have implications for the design of applications to support more effective online searching.

An Algorithm for Efficient Identification of Branched Metabolic Pathways

This article presents a new graph-based algorithm for identifying branched metabolic pathways in multi-genome scale metabolic data. The term branched is used to refer to metabolic pathways between compounds that consist of multiple pathways that interact biochemically. A branched pathway may produce a target compound through a combination of linear pathways that split compounds into smaller ones, work in parallel with many compounds, and join compounds into larger ones. While branched metabolic pathways predominate in metabolic networks, most previous work has focused on identifying linear metabolic pathways. The ability to automatically identify branched pathways is important in applications that require a deeper understanding of metabolism, such as metabolic engineering and drug target identification. The algorithm presented in this article utilizes explicit atom tracking to identify linear metabolic pathways and then merges them together into branched metabolic pathways. We provide results on several well-characterized metabolic pathways that demonstrate that the new merging approach can efficiently find biologically relevant branched metabolic pathways.

Identifying branched metabolic pathways by merging linear metabolic pathways

This paper presents a graph-based algorithm for identifying complex metabolic pathways in multi-genome scale metabolic data. These complex pathways are called branched pathways because they can arrive at a target compound through combinations of pathways that split compounds into smaller ones, work in parallel with many compounds, and join compounds into larger ones. While most previous work has focused on identifying linear metabolic pathways, branched metabolic pathways predominate in metabolic networks. Automatic identification of branched pathways has a number of important applications in areas that require deeper understanding of metabolism, such as metabolic engineering and drug target identification. Our algorithm utilizes explicit atom tracking to identify linear metabolic pathways and then merges them together into branched metabolic pathways. We provide results on two well-characterized metabolic pathways that demonstrate that this new merging approach can efficiently find biologically relevant branched metabolic pathways with complex structures.

Bipolarity of the Saccharomyces Cerevisiae Genome

Accumulating evidence indicates that eukaryotic genes tend to belong in two distinct categories that we will call class I and class II. Class I genes do not contain a TATA box in their promoter, and have low expression variability both at the single cell level (in constant environment) and at the population level (in changing environmental conditions). In contrast, class II genes contain a TATA box in their promoter, and tend to have pronounced expression variability both at the single cell level (in constant environment) and at the population level (in changing environmental conditions). Here we show that the positioning and regulation of class I and class II genes is strikingly different in the large-scale transcriptional regulatory (TR) network of S. cerevisiae. We also show that class I and class II genes differ dramatically in several properties, including gene expression variability at diverse time scales and population sizes, mutational variance, gene essentiality and subcellular localization. This dichotomy might indicate that evolution placed different genes in different locations within the cell and within the TR network, according to some fundamental principles that govern cellular information processing and survival in a changing environment.

Visualizing the Results of Metabolic Pathway Queries

Biology contains a wealth of network data, such as metabolic, transcription, signaling and protein-protein interaction networks. Our research currently focuses on metabolic networks, although similar ideas may be applied to other biological networks. Metabolic networks consist of the chemical compounds and reactions necessary to support life. Traditionally, series of successive metabolic reactions have been organized into simple metabolic pathways and manually drawn. However, as we move into the era of systems biology, it is becoming apparent that automated ways of processing and visualizing metabolic networks must be developed.