John P. Miller - US grants

The goal of this research project is to determine how a relatively complex sensory system extracts and encodes information from external stimuli. The response properties of the system will be observed and characterized at two levels: at the receptor level and at the output layer to higher CNS levels. During some experiments, the responses from all of the cells in the output layer will be monitored simultaneously. Analysis of this data will provide a complete description of the input/output properties and encoding algorithm of this system. The preparation to be studied is the cercal sensory system of the cricket, Acheta domesticus. Crickets (and many other insects) have two antenna-like appendages at the rear of their abdomen, covered with hundreds of "filiform" hairs, resembling bristles on a bottle brush. Deflection of these filiform hairs by wind currents activates mechanosensory receptors, which project into the terminal abdominal ganglion to form a topographic representation (or "map") of "wind space". Primary sensory interneurons having dendritic branches within this afferent map of wind space are selectively activated by wind stimuli with "relevant" parameters, and generate action potentials at frequencies that depend upon the value of those parameters. The "relevant" parameters are thought to be the direction, velocity, and acceleration of wind currents directed at the animal. There are only ten pairs of these interneuons which carry the system's output to higher centers. All ten of these output units are identified, and all will be monitored simultaneously with extracellular electrodes. The following specific questions will be addressed: What are the response properties of th sensory receptors, and what are the I/O properties of the receptor layer as a whole? What are the response properties of all the units in the output layer? Is all of the direction, velocity and acceleration information that is extracted at the receptor layer also available at the output layer? How is that information encoded? Are any higher order "features" also encoded? What is the overall threshold, sensitivity and dynamic range of the system as a whole for detecting features of wind stimuli? An attempt will be made to assess how all of the observed properties relate to the structure of the output neurons, the synaptic connectivity of the output interneurons with other local interneurons, and the synaptic connectivity of the output neurons with the sensory afferents in the map of wind space.

This action is to support a workshop on computational neurobiology. The workshop is to be held in San Francisco, California, July 23-24, 1991. The sessions to be held at the workshop will focus on biological issues in neural networks. Not only will leaders in the field be presenting information relevant to computational issues surrounding biological neural systems but a pre-workshop, tutorial session is planned to allow hands-on experience for mathematicians and theorists as well as for experimentalists. This workshop includes students and an informal atmosphere in which ideas can be discussed.

The overall goal of this research is to determine how the neurons in a relatively complex, topographically-mapped sensory system extract information about stimuli and encode that information in their spike trains. The specific aims of the studies are: 1) to determine what parameters of sensory stimuli are encoded in the spike trains of the sensory receptors and projecting interneurons in this system; 2) to determine the accuracy with which that information is encoded; 3) to determine how the information is encoded within different aspects of the spike train patterns; 4) to examine the mechanisms through which the observed coding scheme is actually implemented within this neural network, 5) to determine the extent to which the observed accuracy approaches the theoretical maximum limits, given the constraints imposed by the "physics" of the stimulus environment, and 6) to examine the principles through which features of the system have been optimized to meet those constraints. The preparation to be studied is the cercal sensory system of the cricket. This system mediates the detection and analysis of low velocity air currents in the animal's immediate environment. All relevant sensory information is carried to higher centers by only ten pairs of primary sensory interneurons. All of these output units are identified, and all will be monitored simultaneously with extracellular electrodes. The first three goals listed above will be achieved by carrying out several types of electrophysiological input/output analyses, primarily at a systems level. Principles of information theory will be applied to the data to obtain quantitative, model-independent measures of the amount of information encoded within the spike trains of sensory receptors and primary sensory interneurons. The fourth goal will be achieved through further electrophysiological measurements, the results of which will be embodied in a physiology-based model of the system. The model will be refined, and its validity tested, on the basis of information theoretic analyses similar to the ones that were carried out on the real system. The fifth goal will be achieved through calculations of the constraints on the system's performance imposed by thermal noise in the air current environment. The sixth goal will be achieved through computer simulations of mechanoreceptor and interneuron characteristics. The studies will elucidate general principles related to optimal signal processing within sensory systems and may suggest general algorithms for efficient coding of information by ensembles of neurons.

The overall goal of this research is to determine how the neurons in a relatively complex, topographically-mapped sensory system extract information about stimuli and encode that information in their spike trains. The specific aims of the studies are: 1) to determine what parameters of sensory stimuli are encoded in the spike trains of the sensory receptors and projecting interneurons in this system; 2) to determine the accuracy with which that information is encoded; 3) to determine how the information is encoded within different aspects of the spike train patterns; 4) to examine the mechanisms through which the observed coding scheme is actually implemented within this neural network, 5) to determine the extent to which the observed accuracy approaches the theoretical maximum limits, given the constraints imposed by the "physics" of the stimulus environment, and 6) to examine the principles through which features of the system have been optimized to meet those constraints. The preparation to be studied is the cercal sensory system of the cricket. This system mediates the detection and analysis of low velocity air currents in the animal's immediate environment. All relevant sensory information is carried to higher centers by only ten pairs of primary sensory interneurons. All of these output units are identified, and all will be monitored simultaneously with extracellular electrodes. The first three goals listed above will be achieved by carrying out several types of electrophysiological input/output analyses, primarily at a systems level. Principles of information theory will be applied to the data to obtain quantitative, model-independent measures of the amount of information encoded within the spike trains of sensory receptors and primary sensory interneurons. The fourth goal will be achieved through further electrophysiological measurements, the results of which will be embodied in a physiology-based model of the system. The model will be refined, and its validity tested, on the basis of information theoretic analyses similar to the ones that were carried out on the real system. The fifth goal will be achieved through calculations of the constraints on the system's performance imposed by thermal noise in the air current environment. The sixth goal will be achieved through computer simulations of mechanoreceptor and interneuron characteristics. The studies will elucidate general principles related to optimal signal processing within sensory systems and may suggest general algorithms for efficient coding of information by ensembles of neurons.

This award will support the development advanced software tools for establishing a database containing information about the structural and dynamic functional attributes of identified neurons, and the development of advanced software tools for querying that database for information about the global functional organization of neuronal ensembles. The visual format of the databases developed with these tools will be multidimensional "atlases", which preserve the spatial relationships between all objects within the nervous systems. These project will be carried out as a collaborative effort between a group of neurobiologists and computer scientists at the Berkeley and San Diego campuses of the University of California. The major specific aims are to develop a general database engine capable of representing information about the functionally relevant neuronal attributes of the objects (such as their biochemical, biophysical, genetic and developmental characteristics); the 3D anatomical parameters of neural structures across multiple resolution scales; the functional characteristics of these objects which determine their individual "input/output" properties; and the functional relations between neuronal objects which determine their dynamic interactions; to develop advanced software tools for inserting the anatomical, functional and relational information into the database and to develop advanced query tools for retrieving and analyzing the data.

The overall goal of this research is to determine how the neurons in a relatively complex, topographically-mapped sensory system extract information about stimuli and encode that information in their spike trains. The specific aims of the studies are: 1) to determine what parameters of sensory stimuli are encoded in the spike trains of the sensory receptors and projecting interneurons in this system; 2) to determine the accuracy with which that information is encoded; 3) to determine how the information is encoded within different aspects of the spike train patterns; 4) to examine the mechanisms through which the observed coding scheme is actually implemented within this neural network, 5) to determine the extent to which the observed accuracy approaches the theoretical maximum limits, given the constraints imposed by the "physics" of the stimulus environment, and 6) to examine the principles through which features of the system have been optimized to meet those constraints. The preparation to be studied is the cercal sensory system of the cricket. This system mediates the detection and analysis of low velocity air currents in the animal's immediate environment. All relevant sensory information is carried to higher centers by only ten pairs of primary sensory interneurons. All of these output units are identified, and all will be monitored simultaneously with extracellular electrodes. The first three goals listed above will be achieved by carrying out several types of electrophysiological input/output analyses, primarily at a systems level. Principles of information theory will be applied to the data to obtain quantitative, model-independent measures of the amount of information encoded within the spike trains of sensory receptors and primary sensory interneurons. The fourth goal will be achieved through further electrophysiological measurements, the results of which will be embodied in a physiology-based model of the system. The model will be refined, and its validity tested, on the basis of information theoretic analyses similar to the ones that were carried out on the real system. The fifth goal will be achieved through calculations of the constraints on the system's performance imposed by thermal noise in the air current environment. The sixth goal will be achieved through computer simulations of mechanoreceptor and interneuron characteristics. The studies will elucidate general principles related to optimal signal processing within sensory systems and may suggest general algorithms for efficient coding of information by ensembles of neurons.

A confocal microscope system will be purchased for studies on such problems as the functional organization of neurons within a sensory system, the developmental mechanisms involved in neuronal and glial differentiation, the nature of the neural code, i.e., the algorithms through which information is encoded in neuronal spike trains, the functional organization of the anterior pituitary, the migration of cells on and through basement membranes in multicellular organisms, interactions between the cell surface of fungi and host tissues and the structural, elemental and molecular preservation of dinosaur tissues.

This award is made under the high performance connections portion of ANIR's "Connections to the Internet" announcement, NSF 96-64. It provides partial support for two years for a DS-3 connection to the vBNS. Applications include projects in computational biology, physics and chemistry. Collaborating institutions include the University of California at San Diego, California Institute of Technology, Courant Institute, University of California at Davis, University of Tennessee, University of California at Santa Cruz and the University of Wisconsin.

06/19/98 This proposal will fund the purchase of a computing cluster to support computational biology research at Montana State University. The computing equipment will be used for multidisciplinary research in the general areas of Neurosciences, Population Ecology, Biochemistry, and Structural Biology. The major users will be 12 faculty from four different departments at Montana State University: Biology, Chemistry and Biochemistry, Mathematics, and Computer Science. A total of 11 postdocs and 27 graduate students in these faculty labs will use the equipment for their research projects. This computer cluster will also directly support research projects involving numerous undergraduate students, including those enrolled in several special programs for minority students. These students will have access to the equipment though their involvement in research projects sponsored by the faculty users.

DESCRIPTION (Adapted from applicant's abstract): The computational Neuroscience Annual Meetings will be held in the 3rd week of July in each of the next 5 years. CNS*99 will be held in Monterey, California, and subsequent meetings will alternate between Boston, Monterey, and Bozeman, Montana. The CNS meetings serve as a format for the presentation and discussion of research that employs theoretical and/or experimental methods to study the functional organization and operation of nervous systems. Presentations at the meetings focus on the nature of the processing tasks or "computations" executed by nervous systems, the codes by which information is represented during the execution of these tasks, and the structure of the neural machinery through which the computational algorithms are implemented. This meeting is intended to facilitate interdisciplinary interactions between experimentalists and theorists using a wide variety of preparations and approaches, and to help those researchers discover and articulate the general principles that emerge from these studies. The meetings are open to all interested registrants. Attendance at past meetings has been approximately equally distributed between graduate students, post-doctoral researchers and senior researchers.

This Integrative Graduate Education and Research Training (IGERT) award supports the establishment of a multidisciplinary graduate training program of education and research on the structure and function of complex biological systems. The program will focus on integrating knowledge and developing models of biological systems across organization levels and at multiple spatial and temporal scales. Systems to studied range from macromolecular complexes to networks of interacting nerve cells. Training will emphasize understanding complex biological systems in terms of the structures and interaction of their components, and will integrate advanced computational and mathematical approaches with a wide variety of experimental and analytical techniques. Key educational aspects are: two new, year-long, multidisciplinary courses for all students; lab rotations with a training focus, seminars in professional development and ethical practices of science, dual advisors, student involvement in collaborative, multidisciplinary research projects, a dedicated research seminar series, and intensive summer workshops led by investigators from other universities and industry. A highly productive group of faculty will participate, collectively spanning eight departments and three colleges of Montana State University-Bozeman. In conjunction with existing, successful MSU programs, special efforts are planned to increase the number of Native American students pursuing doctoral degrees in the biological sciences.

IGERT is an NSF-wide program intended to facilitate the establishment of innovative, research-based graduate programs that will train a diverse group of scientists and engineers to be well-prepared to take advantage of a broad spectrum of career options. IGERT provides doctoral institutions with an opportunity to develop new, well-focussed multidisciplinary graduate programs that transcend organizational boundaries and unite faculty from several departments or institutions to establish a highly interactive, collaborative environment for both training and research. In this second year of the program, awards are being made to twenty-one institutions for programs that collectively span all areas of science and engineering supported by NSF. This specific award is supported by funds from the Directorates for Biological Sciences, for Computer and Information Science and Engineering, and for Education and Human Resources.

DESCRIPTION (taken from application): A summer research experience program is proposed for undergraduates majoring in the physical and computational sciences or engineering, to introduce them to the study of complex biological systems. Students will be recruited to work in the laboratories of ten NIH-funded investigators that are members of an interdisciplinary graduate program in "Structural and Functional Analysis of Complex Biological Systems". This program is designed to train students to work at the interface between biology and the physical and computational sciences, and involves faculty from the Center for Computational Biology and the Departments of Cell Biology & Neuroscience, Chemistry & Biochemistry, Computer Science, Electrical Engineering, Mathematics, Microbiology, and Veterinary Molecular Biology. This program was recently awarded a five-year NSF Interdisciplinary Graduate Education and Research Training (IGERT) graduate training grant. The undergraduate research program proposed here has been designed to integrate with the graduate-training program. Undergraduates will have graduate student mentors from the program, in addition to their faculty mentors. They will also participate in the summer symposia and workshops organized and funded through the NSF-IGERT program, where they will be exposed to distinguished scientists from around the country who are at the forefront of applying computational and systems approaches to biological problems. In addition, there will be weekly seminars and presentations, where the undergraduates will interact with a cohesive group of investigators at MSU that are committed to training students to exploit advanced experimental and computational techniques to understand complex biological systems. By systems, we mean groups of interacting biological components, such as enzymes, substrates, and regulatory agents in a biochemical pathway, macromolecular assemblies of a second-messenger signaling system, gene products and control elements in a genetic network, or nerve cells in a sensory network. By necessity, the education and training of such scientists must be multidisciplinary and will involve computational approaches and bioinformatics. This presents an excellent opportunity for undergraduates with a background in physics, mathematics, computer science, or engineering to apply their quantitative and analytical skills to exciting new problems of biological or medical interest, and develop an interest in pursuing interdisciplinary graduate training.

A fundamental question in neuroscience is how natural sensory stimuli are encoded for information handling by the brain. Invertebrate animals often offer systems that are in some ways simpler than those of mammals, and including such features as identifiable single cells in networks of relatively few numbers. This collaborative project exploits a sensory system called the cercal system of the cricket, in which small appendages on the rear of the body contain fine hairs that are used to detect, identify and localize behaviorally relevant air current movements, such as those produced by a predator. The input from roughly 2000 receptor cells converges on 30 local interneurons and only 20 output interneurons that lead to behavior such as escape. Three collaborators at two institutions use computational and mathematical analyses of a database of anatomical and physiological measurements on the 'dynamic map' that does the central processing in the brain of the peripheral signals. The goals are to characterize the representation of dynamic sensory stimulus parameters at two processing stages within the mapped sensory system, and to examine the mechanisms that transform the representation at the interface between these two processing stages.
Results will be important for our understanding of information representation in nervous systems, particularly in dynamic processing. The project also will enhance the independent career of a woman faculty member in mathematics, and students will receive multi-disciplinary, highly quantitative training related to biology, in two states that do not currently have high profiles in federally funded research.

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.

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.

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.

DESCRIPTION (provided by applicant): Science Montana: Engaging 4-H Teens with Bioscience Research Many Montana high school students, particularly those in rural and often isolated areas, have limited, if any, exposure to scientists and thus often lack the awareness and the role models needed to consider bioscience study and careers. The proposed project--Science Montana: Engaging 4-H Teens with Bioscience Research--will engage rural teens, address future scientific workforce needs, and leverage the extensive expertise of the state's scientists. It is an innovative partnership of three Montana State University-Bozeman (MSU) entities: the Department of Cell Biology and Neuroscience, Extended University, and Montana 4-H within the Extension Service. In addition, the project collaborates with MSU's Howard Hughes Medical Institute Undergraduate Biology Program and the NCRR-funded INBRE program. The project builds on the resources and infrastructure of the Montana 4-H program with 4-H clubs located throughout the state including Montana's seven Native American reservations. Drawing upon the research interests of MSU science faculty, the project will introduce rural teen participants to basic, applied, and translational research themes using neuroscience (basic), infectious disease (applied), and metabolomics (translational) as instructional examples. The proposed project focuses on an intensive on-campus experience followed by year-long inquiry-based activities and interaction with scientists and student mentors. It will utilize research projects, monthly videoconference lab meetings, online interactive media, and online communications to engage the 4-H participants. The project aims are: 1) engage rural Montana teens in basic, applied, and translational research in biosciences through a year-long inquiry-based learning environment; 2) Increase the interest of rural Montana teens in pursuing basic and clinical bioscience research and other careers related to health sciences, particularly among populations underrepresented in bioscience professions; and 3) create a web-based repository of project content and interactive multimedia assets that will enable 4-H leaders and informal education professionals to incorporate bioscience content, activities, and resources at their local level. Participants will have the opportunity to work with INBRE faculty located at Tribal Colleges across the state and will make public presentations at local community events, state 4-H Congress, and other regional and national meetings.