Tomas Gedeon - US grants

EIA-0129895 -John P. Miller-Montana State University-Algorithms for real-time decoding and modulation of neural spike trains-A grand challenge in neuroscience is to understand the biological basis of information processing in nervous systems. Three major goals facing sensory neuroscientists are a) to understand how sensory information is encoded in the activity patterns of neural ensembles, b) to understand how those activity patterns are decoded by cells at the subsequent processing stages, and c) to understand how computations (e.g. visual pattern recognition or oriented motor responses) are carried out on that decoded information. Two major goals of the research proposed here are a) to develop a formal, general approach toward achieving those goals, and b) to test and refine that approach by characterizing the functional organization and neural encoding scheme of a simple sensory system. These goals will be achieved through the development of a data-driven model of the system. The model will be formulated in terms of information processing units and information channels, rather than in terms of individual neurons. That is, the functional units in the model will be operators that carry out specific, independent computations (or information transformation operations) at a specific processing stage in the test nervous system, and the channels through which information is passed between these functional units will correspond to information channels in the Shannon sense.

Using a combinatorial decomposition of a genetic regulatory network into sub-networks and harnessing the versatility of yeast to encode each of these as a separate strain, this research will dissect, deconvolve and decipher the dynamics of nitrogen catabolite repression. Based on detailed measurements using transcription factor fusion proteins, live cell imaging, and quantitative measurement of transcription rate, the investigors will develop precise models of the dynamical behavior of the complex yeast network. Quantitative models will provide a platform for for the discovery of mathematical theorems, that relate the structure of genetic circuits to their dynamics and function. This theory in turn will allow a study of the robustness of the circuit with respect to parameters and its output. The mathematical theory that will be developed will, among other things, allow researchers to streamline simulations of large biological networks.

In recent years there has been a growing awareness of the connections between biological organisms and engineering systems. There are networks of genes that control and regulate the behavior of other genes and ultimately all biological processes. These networks of genes are analogous to electrical circuits and switches. There has been an expanding willingness and desire to use mathematical and computational techniques to leverage all existing information regarding the structure of biological control circuitry to understand the basic rules on which they are predicated and to acquire the ability to predict their output. Understanding the dynamics of such networks promises revolutionary changes in the treatment of human diseases as well as providing insight for the development of a new generations of robust electronic and digital devices. This project will have a significant impact on graduate education and on recruitment of mathematics graduate students for interdisciplinary scientific work as well as the recruitment of biology, computer science and medical students into the interdisciplinary and emerging field of biomedical informatics.

An implicit hypothesis underlying much recent research in neuroscience and neuroethology is that sensory systems have evolved, through natural selection, toward optimal functional performance and/or energetic efficiency. However, it has proven extremely difficult to derive precise definitions for functional optimality and efficiency, and even more difficult to determine the nature and relative importance of different factors that might be constraining this process of optimization. A multidisciplinary group of researchers lead by Dr. Gedeon will develop a theoretical framework for defining and assessing optimality of one specific sensory system and are also carrying out experiments to assess its optimality and efficiency.
The system they are studying is the cercal sensory system of the cricket, Acheta domesticus. This system functions as a low-frequency, near-field extension of the animal's auditory system, and mediates the detection, localization and identification of signals generated by predators, mates and competitors. The sense organ for this system consists of a pair of antenna-like 'cerci' at the rear of the cricket's body, each of which is covered with approximately 1000 mechanosensory hairs. Each of these hairs is attached to a single nerve cell. The group's working hypothesis is that the biomechanical and neurophysiological characteristics of these receptor organs are optimized for the sensory processing operations they mediate. The researchers will determine the extent and nature of optimization in the array of mechanosensory hairs and receptors, and will also identify specific constraints under which the optimization has taken place. For example, they will determine whether the physical structures of the hairs are matched to behaviorally relevant air-current signals and also determine if the global configuration of the ensemble of hairs on the two cerci reflects optimization with respect to sensitivity, robustness to noise, and/or to the detection of specific types of signals having particular behavioral importance, such as those from predators. They will also characterize constraints on optimization related to biomechanics, resource utilization, and efficiency of subsequent processing operations. These aims are being achieved through a combination of mathematical analysis, computer simulation, quantitative morphometric analysis of the sensory structures, and neurophysiological experiments.
Graduate students in Mathematics and Neuroscience will be involved in the project, and an interdisciplinary graduate-level course is being developed that focuses on optimality in neural systems. Further, in collaboration with the American Indian Research Opportunities program at Montana State University, Native American students at the undergraduate and pre-college levels will carry out many of the experiments and associated data analysis.

This project concerns the development of mathematical models of biochemical oscillations and the analysis required to understand their underlying behavior. While the list of molecular systems that exhibit oscillations is growing, the grasp of the common principles is challenged by the variety of underlying network structures. The presence of negative feedback is the only feature shared by all oscillatory networks; beyond this each system seems to be unique. The goal is to formulate a common principle that is shared by all oscillatory networks. It is expressed in terms of broad tendencies of the dynamics, rather then the detailed structure of the supporting networks. Cells constantly communicate with their environment and their neighbors. This mutual communication often results in synchronization of their behavior. A theory will be developed to explain synchronization of periodic signals in cell communities. The synchronization of biochemical oscillators will be studied via the nearest neighbor coupling, and the coupling with a communally produced and sensed molecule.

A recent unprecedented explosion in our knowledge of structure and function of genetic regulatory networks promises revolutionary changes in how human disease is viewed and treated. Many biological processes ranging from the cell cycle to the human 24 hour circadian rhythm are periodic in time. This periodicity is driven by oscillations in the concentration of chemicals in individual cells. In many cases these signals synchronize across populations and the periodic signal can be detected on a macroscopic level. The regulatory networks that are responsible for production and maintenance of these rhythms are being discovered every day. While the structures of these networks vary, they often produce similar periodic signals. This project will study the underlying broad dynamics principles behind the oscillations, emphasizing the function of the network over its implementation. Further, it will study synchronization of the periodic rhythms in population of cells. The lack of synchronization in the population can result in developmental and functional defects. This project will develop a theory that will be able to predict collective dynamics, as well as suggest repair mechanisms in the case of failure.

The ability to identify micro-flow characteristics using small sensors (1 mm or less) is becoming increasingly important in many engineering applications. In biomedical engineering there is a need to measure local flow properties in blood vessels because these fluid properties have a potential impact on the structural integrity of the vessel wall. In aerospace engineering, micro-planes are being developed for a number of applications, but the performance of these micro-planes is limited due to the difficulties of preventing flow separation along the wings and the resulting stall. Measuring these characteristics while the micro-plane is in flight is proving to be a significant challenge. While engineers have been grappling with the design of micro-flow-sensors for a few decades, crickets and other arthropods have used a few million years of evolution to develop micro-flow-sensors that are essential for threat detection, predator avoidance, and communication. In the common house cricket the micro-flow-sensors are two antenna-like appendages, called cerci, at the rear of the abdomen. Each cercus is covered with approximately 800 filiform mechanosensory hairs, each of which is connected to a single spike-generating neuron. Deflection of a hair by air currents changes the spiking activity of the associated receptor neuron at the base of the hair. It has been shown that the cercal system is extraordinarily sensitive and capable of detection of air motion caused by thermal noise. This sensitivity is beyond the capability of current artificial micro-flow sensors. Our project is based on the hypothesis that a better understanding of the arthropod micro-flow-sensor can guide the development and improvement of artificial micro-flow-sensors. We will develop new modeling and computational tools for the unsteady Stokes equations based on immersed boundary techniques to study the cercal micro-sensor in crickets. These models will be directly applicable to artificial micro-flow sensors.

It has been recognized for many years that engineering design can greatly benefit from the understanding of biological structures. Evolution has resulted in very sophisticated solutions for complex problems related to the detection and analysis of very small air and fluid movements in an animal's immediate environment, and aspects of those biological solutions should be directly applicable or generalizable, in principle, for engineered systems. An interdisciplinary team that combines an engineer, mathematician and a neurobiologist will develop new modeling and computational tools to study performance characteristics of the cercal micro-flow sensor in crickets. Two principal outcomes will be directly applicable to design of artificial micro-flow sensors. The first outcome will be a collection of computational techniques and models which explicitly address the effect of fluid motion on the sensors. The second outcome will be a set of biological principles that evolved in response to constrains posed by the physics of structure-fluid interactions, and that can guide the development of artificial flow sensors.

In the past twenty years biology has greatly benefited from insights gathered from mathematical models of biological processes. The unprecedented growth of molecular and systems biology in last 10 years challenges mathematicians to develop modeling techniques appropriate for the intricate biology on a cellular and sub-cellular level. It is increasingly clear that the processes in a living cell cannot be captured by static descriptions, but must be modeled as dynamical processes. This leads to descriptions by sets of differential equations, often with a stochastic component and including spatial dependencies. The conference will bring together mathematical biologists working on systems ranging from individual molecular processes to entire ecosystems. The aim of the conference is to facilitate the discovery of similarities between systems at these widely different scales.

The revolution in genetic and molecular biology of the past decade promises a great advancement in medical science in the near future. The progress on many human diseases like cancer, Alzheimer's disease and genetic diseases is tied to our understanding of molecular mechanisms causing them. Every second of every day hundreds of chemical reactions in each cell sustain our lives. While we increasingly know who the players in these reactions are, we do not understand how they are controlled and regulated. Mathematical models can guide the necessary experiments that are needed to untangle this clockwork of interconnected processes. The conference will bring together the leading scientists in mathematical systems biology to exchange the ideas and establish new collaborations in order to help molecular biology to deliver on its medical promise. The conference participants constitute a diverse group of speakers in terms of gender, geographic origin, as well as career stages, and include members of underrepresented groups in science.

DESCRIPTION (provided by applicant): We developed systems to investigate hepatocyte lineage life history dynamics in vivo. We propose to define the factors that determine hepatocyte lineage birth-rates and longevities, and to describe their dynamic responses to hepatic stresses in aging. In this collaborative proposal, empirical in vivo studies are combined with mathematical modeling and simulation to test effects of extrinsic and intrinsic factors on hepatocellular lineage dynamics and how these change as a part of the aging process in a genetically tractable animal model. What is known. Unlike most differentiated cell types, hepatocytes can proliferate. When normal liver cells are transplanted into mice having a genetic defect that autonomously compromises the endogenous hepatocytes, the grafted cells can complete >18 consecutive replicative cycles, resulting in full replacement of the endogenous hepatocytes and reconstitution of the liver with healthy cells. Using serial reconstitution through 7 consecutive recipient mice, a classic study showed that adult wild-type liver cells could undergo an average of at least 69 consecutive divisions 1. Thus, liver cells have a nearly unlimited capacity to proliferate 1 and may be "stem cell-like" in their regenerative immortality 1,2. Hepatocytes are also one of few cell types that undergo endoreplication and acytokinetic mitosis, resulting in polyploid nuclei and bi-nucleate cells, respectively. Most hepatocytes are polyploid and both diploid and polyploid hepatocytes can proliferate. Indeed, another study showed that the most active liver cell types for reconstituting compromised liver are polyploid hepatocytes 3. These background data indicate that: (1) liver cell populations may be infinitely proliferative;(2) hepatocytes are predisposed to becoming polyploid;and (3) polyploid hepatocytes are highly proliferative. Preliminary observations and the problem they reveal. We developed genetic marker systems for "time-stamping" hepatocyte lineages in vivo 4. In contrast to the prevailing model of the immortal hepatocyte, our systems show that hepatocyte lineages have both a finite half-life and a limited capacity to proliferate. We also developed a flow cytometry-based method of quantifying liver nuclei on the basis of ploidy and found that adult livers between 2- and 12-months of age exhibited nearly invariant ratios of diploid (2N), 4N, and 8N nuclei. Lastly, we developed a novel "ten-day chronometer" for newly differentiated hepatocyte lineages that allows us to quantitatively assess the contributions of pre-hepatocytic stem cells to liver growth, regeneration, and maintenance 5. Using this chronometer, we found that normal adult liver is continuously gaining new diploid hepatocyte lineages. We believe these replace lineages that die-off due to age or stress. Based on our observations, we suspect that only pre-hepatocyte cell types, not differentiated hepatocytes, have unlimited proliferative potential and that this rare population of cells underlies the proliferative immortality of liver. Our hypothesis and how we will test it. Based on our findings, we hypothesize that hepatocytes have a life history that includes birth from stem cells, age-related deterioration, and death. We predict that hepatic stresses, replication, ploidy, aneuploidy, and time will affect hepatocyte lineage life history dynamics. Moreover, the process of aging of the host animal might influence the life history dynamics of hepatocyte lineages. The quality of either the hepatic stem cells or the "liver niche" could change as livers age, resulting in differences in hepatocyte lineage birth rates, longevities, proliferative potentials, and stress resistance. To test our hypothesis, we will fulfill four aims: (1) Define birth-rates and longevities of hepatocyte lineages under normal and stressed conditions. (2) Determine what factors limit lineage longevity. (3) Measure the dynamic lineage-aging process in hepatocyte nuclei. (4) Examine how an animal's age and exposure-history affects the life history dynamics of hepatocyte lineages. Implications for human health in aging. Hepatocytes are generally thought to have a stem cell-like capacity to proliferate and regenerate lost or damaged liver tissue. Indeed, the term 'stem cell-like'invokes a level of immortality that has been tested in only a small number of situations. Clearly mouse ES cells, for example those we used to make our various lines of mice having targeted mutations, have been verified through years of culture and mouse-production as being indefinitely self-renewing;but is this true of all organ-specific stem cells? Maintenance of the self-renewing capacity of ES cells in culture requires a very strict environment or "niche" (e.g., media, supplements, attachment factors, feeder-cells, pH, etc.), so it may be reasonable to predict that the "quality" of organ-specific stem cells in vivo could also be intimately dependent on the niche that the host-organ provides. This niche could change with an animal's age or exposure history, yet these possibilities have not been previously considered. Here we propose an investigation into the aging process in differentiated hepatocyte lineages and how both this process and the contributions of hepatic stem cells change as animals age. Upon completion of this project, we will have: (a) developed and publicly disseminated novel mouse models for studying aging of hepatocyte lineages;(b) defined rates of birth and death of hepatocyte lineages under normal and several stressed states;(c) characterized the aging process in normal and stressed hepatocyte lineages;and (d) investigated how these processes change as a function of aging and exposure history of the host animal. PUBLIC HEALTH RELEVANCE: Mice, like humans and likely all mammals, undergo a process of age-dependent deterioration, or "aging", which is associated with accumulation of molecular damage and often with a reduced capacity to renew or regenerate existing tissue. Many adult tissues have tissue-specific stem cells, which are thought to possibly play a role in rejuvenating the tissue and thereby countering or delaying the aging process;however in most cases, these cells are rare and their activities are poorly studied. Liver is a large metabolically active organ that plays numerous crucial roles in organismal physiology. The organ is composed predominantly of a single cell type, the differentiated hepatocyte, which is responsible for most of the liver's functions. Unlike most differentiated cells in the body, hepatocytes can proliferate and self- renew. Indeed, currently accepted models suggest that hepatocytes are immortally self-renewing, and therefore play a stem cell-like role in organ maintenance. Curiously, it is known that liver does contain a small population of hepatic stem cells;however, no role for these cells in normal liver physiology or maintenance has yet been established. It is our belief that some of these properties of hepatocytes and hepatic stem cells are inaccurate, and stem from a lack of suitably sensitive means of assessing either the immortality of hepatocyte lineages or the contributions of hepatic stem cells. We have developed innovative and highly sensitive fluorescent lineage-marker systems for measuring rates of birth of new hepatocyte lineages from pre-hepatocytes and for following hepatocyte lineages and quantitatively assessing their decay in vivo. As such, these systems allow very sensitive measures of processes that had previously been undetectable, and they are revealing some unexpected properties of hepatocytes and hepatic stem cells. Our preliminary studies show that pre-hepatocytes, most likely the hepatic stem cells, play a constant role in generating new lineages of proliferative hepatocytes, and that these lineages decay with time. Thus, we have found that hepatocyte lineages have a life history that includes birth from a pre-hepatocyte "stem cell" population, aging, and death, which is constantly occurring in resting liver. Moreover, we have found that some changes in the physiological state of the liver can alter the rate of birth of new hepatocyte lineages. We believe that this birth of new hepatocyte lineages is of crucial importance for long-term maintenance of the liver. In the proposed project, we test how this contribution changes as an animal ages, how the longevity of hepatocyte lineages change as an animal ages, and whether the "hepatotoxic exposure" or "hepatic stress" histories of an animal alters hepatocyte lineage life history dynamics. The proposed study integrates descriptive studies on hepatocyte lineage dynamics in aging animals with mechanistic studies on the aging process and with computational modeling of the cellular dynamics in young and aged liver. This is a collaborative proposal that integrates the expertise of a mouse molecular geneticist/cell biologist with that of a bio-mathematician to provide an accurate, quantitative, and predictive analysis of the aging process in liver. A plan for integrating the "empirical" and "computational" subgroups of this project is provided. Novel mouse systems are developed within this proposal and a resource-sharing plan is included for these are to be made freely available to the international research community for continued investigations along these or other lines.

DESCRIPTION (provided by applicant): We developed systems to investigate hepatocyte lineage life history dynamics in vivo. We propose to define the factors that determine hepatocyte lineage birth-rates and longevities, and to describe their dynamic responses to hepatic stresses in aging. In this collaborative proposal, empirical in vivo studies are combined with mathematical modeling and simulation to test effects of extrinsic and intrinsic factors on hepatocellular lineage dynamics and how these change as a part of the aging process in a genetically tractable animal model. What is known. Unlike most differentiated cell types, hepatocytes can proliferate. When normal liver cells are transplanted into mice having a genetic defect that autonomously compromises the endogenous hepatocytes, the grafted cells can complete >18 consecutive replicative cycles, resulting in full replacement of the endogenous hepatocytes and reconstitution of the liver with healthy cells. Using serial reconstitution through 7 consecutive recipient mice, a classic study showed that adult wild-type liver cells could undergo an average of at least 69 consecutive divisions 1. Thus, liver cells have a nearly unlimited capacity to proliferate 1 and may be stem cell-like in their regenerative immortality 1,2. Hepatocytes are also one of few cell types that undergo endoreplication and acytokinetic mitosis, resulting in polyploid nuclei and bi-nucleate cells, respectively. Most hepatocytes are polyploid and both diploid and polyploid hepatocytes can proliferate. Indeed, another study showed that the most active liver cell types for reconstituting compromised liver are polyploid hepatocytes 3. These background data indicate that: (1) liver cell populations may be infinitely proliferative; (2) hepatocytes are predisposed to becoming polyploid; and (3) polyploid hepatocytes are highly proliferative. Preliminary observations and the problem they reveal. We developed genetic marker systems for time-stamping hepatocyte lineages in vivo 4. In contrast to the prevailing model of the immortal hepatocyte, our systems show that hepatocyte lineages have both a finite half-life and a limited capacity to proliferate. We also developed a flow cytometry-based method of quantifying liver nuclei on the basis of ploidy and found that adult livers between 2- and 12-months of age exhibited nearly invariant ratios of diploid (2N), 4N, and 8N nuclei. Lastly, we developed a novel ten-day chronometer for newly differentiated hepatocyte lineages that allows us to quantitatively assess the contributions of pre-hepatocytic stem cells to liver growth, regeneration, and maintenance 5. Using this chronometer, we found that normal adult liver is continuously gaining new diploid hepatocyte lineages. We believe these replace lineages that die-off due to age or stress. Based on our observations, we suspect that only pre-hepatocyte cell types, not differentiated hepatocytes, have unlimited proliferative potential and that this rare population of cells underlies the proliferative immortality of liver. Our hypothesis and how we will test it. Based on our findings, we hypothesize that hepatocytes have a life history that includes birth from stem cells, age-related deterioration, and death. We predict that hepatic stresses, replication, ploidy, aneuploidy, and time will affect hepatocyte lineage life history dynamics. Moreover, the process of aging of the host animal might influence the life history dynamics of hepatocyte lineages. The quality of either the hepatic stem cells or the liver niche could change as livers age, resulting in differences in hepatocyte lineage birth rates, longevities, proliferative potentials, and stress resistance. To test our hypothesis, we will fulfill four aims: (1) Define birth-rates and longevities of hepatocyte lineages under normal and stressed conditions. (2) Determine what factors limit lineage longevity. (3) Measure the dynamic lineage-aging process in hepatocyte nuclei. (4) Examine how an animal's age and exposure-history affects the life history dynamics of hepatocyte lineages. Implications for human health in aging. Hepatocytes are generally thought to have a stem cell-like capacity to proliferate and regenerate lost or damaged liver tissue. Indeed, the term 'stem cell-like' invokes a level of immortality that has been tested in only a small number of situations. Clearly mouse ES cells, for example those we used to make our various lines of mice having targeted mutations, have been verified through years of culture and mouse-production as being indefinitely self-renewing; but is this true of all organ-specific stem cells? Maintenance of the self-renewing capacity of ES cells in culture requires a very strict environment or niche (e.g., media, supplements, attachment factors, feeder-cells, pH, etc.), so it may be reasonable to predict that the quality of organ-specific stem cells in vivo could also be intimately dependent on the niche that the host-organ provides. This niche could change with an animal's age or exposure history, yet these possibilities have not been previously considered. Here we propose an investigation into the aging process in differentiated hepatocyte lineages and how both this process and the contributions of hepatic stem cells change as animals age. Upon completion of this project, we will have: (a) developed and publicly disseminated novel mouse models for studying aging of hepatocyte lineages; (b) defined rates of birth and death of hepatocyte lineages under normal and several stressed states; (c) characterized the aging process in normal and stressed hepatocyte lineages; and (d) investigated how these processes change as a function of aging and exposure history of the host animal.

Observations from recent biological experiments indicate that the motion of a transcription complex or a ribosome along the template is subject to multiple pauses of varying time durations. In the case of transcription of ribosomal RNA in bacteria, the density of elongating complexes is sufficiently high to experience traffic jams. Since the growth rate in a fast growing E. coli population is determined by the availability of ribosomes, which in turn is limited by ribosomal RNA transcription rates, the elongation process of ribosomal RNA is a key factor determining the E. coli growth rate. This research project uses a nonlinear hyperbolic partial differential equation (PDE) as a mathematical model for the transcription process of ribosomal RNA. The PDE takes the form of a nonlinear conservation law, and it was first proposed as a mathematical model for traffic patterns in the 1950's. Although this type of model was superseded in the study of traffic flow by models that take into account driver reactions, it is suitable for modeling elongation processes as it incorporates density dependent velocities and the presence of transcriptional pauses. The PDE model is used to estimate an instantaneous elongation rate in the presence of a non-uniform distribution of pauses, which are biologically important yet experimentally inaccessible quantities. The second major element of the research is the development of analytical as well as sophisticated numerical methods for verification, validation and sensitivity analysis of the model. Discontinuous Galerkin methods form the foundation of the computational schemes, allowing efficient numerical solution of systems of nonlinear conservation laws with discontinuous coefficients that encode the presence of pauses. These new tools will be used to determine the limits that transcriptional pauses impose on the mean and variance of the transcription rate of ribosomal RNA.

Maintenance of homeostasis, capacity to grow and ability to respond to stimuli are just a few of the characteristics of life that depend on each cell's ability to regulate production of molecules such as mRNA and proteins. Proteins are produced by ribosomes sliding along a long thin strand of messenger RNA (mRNA). In turn, mRNA is generated by a transcription complex sliding along the DNA template. These translation and transcription processes fall within the broad classification of "elongation processes." Although recent revolutionary advances in cell biology have improved scientists' qualitative understanding of life on a molecular level, efforts to construct predictive quantitative models are often limited by a lack of quantitative experimental data. However, one important system, transcription of ribosomal RNA, is sufficiently experimentally characterized to allow delineation of the limits it imposes on cellular physiology. The current work develops quantitative models of this transcription process in order to quantify limitations it imposes on bacterial growth rates. Connections between the molecular biophysics of elongation processes and the macroscopically observable growth rate are quantitatively studied. The current research combines careful modeling of a complex biological process with the development of new mathematical and numerical techniques to answer fundamental biological questions. The results may have a broad impact on our understanding of bacterial physiology and, through a better understanding of the driving mechanisms of bacterial growth, may have a long term impact on molecular biotechnology.

Microbiology research is currently undergoing a revolutionary transition from the study of primarily monocultures to the study of natural and synthetic microbial communities. Microbial consortia play a key role in chronic medical infections and the human gut microbiome and have been implicated in many chronic medical conditions including ulcerative colitus. Microbial consortia are also used in water treatment plants, toxic site remediation and biofuel production. Understanding the behavior of microbial communities in response to a disturbance and their subsequent control remains an outstanding scientific challenge. This project will provide a first step toward the goal of developing analytic and modeling techniques that will allow rational design of synthetic microbial communities that are robust and controllable.

There are theories suggesting that diversification in a microbial community may enhance stability, robustness and increase efficiency; however a precise quantitative understanding of the advantages that a consortium provides for its members is needed. This project seeks to experimentally and theoretically determine the impact of specific metabolic interactions within a microbial community on emergent properties like enhanced stress tolerance or increased biomass production. The project will use tools of dynamical systems, like invariant manifold theory and stability of invariant sets, to analyze models of microbial communities. Methods of partial differential equations will be used to extend this analysis to interacting spatially segregated communities. These models will precisely quantify the relationship between the number of defined microbial interactions and the magnitude of improvement on both consortia efficiency and enhanced tolerance. This will provide a systems biology basis for characterizing consortial behaviors and a scalable methodology to compare the benefits of membership in a microbial community to the costs for each individual. This information is fundamental to understanding microbial community persistence and will provide a rational framework for controlling both unwanted and desirable consortia. Prior research has yielded preliminary results on the conditions necessary for increased biomass production in specialized binary communities versus a monoculture and has demonstrated the existence of invariant manifolds that simplify the analysis for many types of microbial communities. This project extends this work by building experimental synthetic communities with well-defined metabolic roles, measuring the synthetic community's characteristics like biomass production and robustness to perturbations and constructing mathematical models for more complex and spatially segregated communities.

Molecular and cellular biology are at the forefront of our quest to understand life and improve human health. However, the details of how genes and signals interact in real time to produce cell behavior, are not well understood. The problem is a lack of approaches that can integrate experimental data into models that describe the important behavior of the system, yet do not describe nonessential details, that would make it too cumbersome to compute in a reasonable time. This project will develop a new set of tools, that can predict the entire range of behaviors of a network of interdependent genes and do it so efficiently that behavior under the effect of a few mutations or behavior of alternative sets of genes can be readily explored. Furthermore, once such a description of behavior of a set of genes is available, it will be shared and used by other scientists. Continuous build-up of these results will rapidly increase their value for data science, cellular biology, and broader science.

Molecular and cellular biology have witnessed a huge leap forward since science has acquired the ability to sequence genomes. However, while genomes are relatively static, the phenotypes result from the interaction of many genes that are expressed dynamically in time. Furthermore, since phenotypes arise from complex transcriptional networks, where the interactions tend to be nonlinear, data-based models will play a key role in understanding and ultimately controlling these phenotypes. Current modeling techniques, motivated by physics, struggle to bridge a fundamental conflict between low resolution of biological measurements informing model parameter values and the fundamental fact that dynamical systems are sensitive to initial conditions and parameters. This project develops a novel mathematical framework that provides a quantitative description of global dynamics that are compatible with coarse, noisy biological measurements. Combined with computationally efficient algorithms for the construction of databases that capture biologically relevant dynamics of a network, it can become the basis for shared science in the space of gene networks. These tools will be used to construct and validate network models from experimental data, interrogate dynamic behavior of a given network, and compare dynamics summaries across networks.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.