Lawrence H. Snyder - US grants

A chaotic router is a non-minimal, adaptive message router that uses randomization in routing and derouting decisions. The use of randomization simplifies the router sufficiently that it's conjectured chaotic routing can be competitive with oblivious routing, achieving equivalent or better average-case performance, greatly improved worst-case performance and fault tolerance. Though the principles of chaotic routing apply to some degree to any topology with multiple paths between nodes, chaotic routers for binary n-cubes are presently under investigation. The goal of the research is to establish a compelling case for constructing an experimental version of a hypercube chaotic router. Towards that end a series of studies are presently underway. 1. The idealized router of the proposal is being converted to practical engineering designs; the details of the chaotic routing logic are being specified. Alternatives include sequential and parallel channel match algorithms, a virtual cut-through alternative to the store-and- forward of the proposal, various alternates for asynchronous channels, various queue sizes, etc. 2. A detailed simulator (with a granularity of approximately a half-dozen gate delays) has been created. Experiments are being run on a 256 node hypercube to compare the different chaotic router designs with various alternative routers, such as oblivious and priority routers. 3. Simulations are being run to analyze the effect of multiqueue size on the latency and throughput of the network. The role of message length on latency and throughput is also being analyzed. 4. Probabilistic analysis, completed last summer, shows that an idealized hypercube chaotic router is deterministically deadlock free and probabilistically livelock and starvation free. The new engineering designs are being abstracted and analyzed to assure that they are consistent with the idealized model of this theory. 5. The first version of a custom CMOS channel design is back from fabrication and is being tested. 6. Alternate implementation strategies, besides custom CMOS, are being investigated. 7. The generality of the chaotic routing principles are being analyzed in the context of different topologies. In particular, various forms of the mesh topology are being studied as alternates. 8. The role of clocked and self-timed logic in chaotic routing is being analyzed in light of the fact theoretical foundations are independent of the timing policy.

Three CAD design tools, produced in research projects at the University of Washington, are being developed and enhanced so as to bring them into a state for distribution to general users. These tools are: (1) MacTester, an interactive testing and debugging environment built around the Mackintosh computer; (2) WireC, a mixed graphical and procedural language for describing hardware systems; and (3) Gemini, a VLSI layout verification program that compares the circuit specification with the circuit layout. This grant is made under the Software Capitalization Program.

This award is to support a postdoctoral associate to work in experimental computer science. The associate, Calvin P. Lin, will work with Dr. Lawrence Snyder on portable parallel programs. The research will concentrate on three areas: parallel languages, parallel algorithms, and portable scientific applications. Two parallel languages will be developed based on the Phase Abstractions programming model. This model has already been shown to produce portable parallel programs for both shared and distributed memory architectures. Additionally, important scientific algorithms, such as the Fast Fourier Transform, will be studied to identify the "algorithm of choice", an algorithm that is optimal on most architectures and competitive on all. Finally, work will continue on the development of portable scientific codes concentrating on a large molecular dynamics program.

Snyder The chaotic router for multiprocessor systems avoids congestion in message routing by derouting packets chosen at random at congested nodes of a network. The routers can thus adapt to varying message traffic. In this project, the router is being implemented and its performance is being measured within a testbed that approximates a real multiprocessor.

This is a conference that draws together researchers from both universities and industry across a wide range of disciplines. The emphasis of the conference is the design of systems taking into account chip design, packaging, and interconnection issues. By holding one session track instead of parallel sessions, the organizers emphasize the importance of a firm grasp of all of these issues for the modern system designer.

The theme of our program is particle simulation. In order to solve physical problems, we will include: The computer science, applied math and physical science necessary to perform state-of-the-art simulations on massively parallel computers. The creation of tools to use parallel computers to drive visualization engines. Integration of simulation with experimentation. For large experimental projects, simulations are key to their planning and interpretation. By concentrating on a single computational theme, we will build a cohesive program spanning several departments. The training program will be jointly administered by Astronomy, Chemistry and Computer Science. Our program has the following key features: It is a cohesive multidisciplinary program with a formal mechanism of joint supervision of students between departments. We will train scientists and engineers in the use of high performance computers for simulation, analysis and visualization. We have arranged summer internships at government labs and in industry. We propose a specific program of action to improve the number of women and minorities in our program until it reflects the diversity of the US population.

This research project is directed at the design and detailed specification of the Advanced ZPL language in anticipation of its full experimental evaluation. Advanced ZPL is intended as the programming language for scientific and engineering computations that are to be run on parallel supercomputers. Unlike ZPL (and other antecedent data parallel languages) Advanced ZPL is suitable for general parallel algorithms. Advanced ZPL contains ZPL as a proper subset, giving it full support for data parallel computation. In addition, Advanced ZPL provides the capability to express thread parallelism, task parallelism and pipelining. ZPL provides array data structures, while Advanced ZPL adds general data structuring capabilities. Control over data motion and computational granularity are provided. The goal is to produce a precise specification with which to write sample applications. Based on these applications users can assess the convenience and expressivity of Advanced ZPL, and compiler writers can formulate ways to produce efficient, portable code for it.

DESCRIPTION (Adapted From The Applicant's Abstract): This proposal will investigate the role of the posterior parietal cortex in directing movements to visible targets. The parietal cortex clearly plays a role in this process, but the nature of that role is unresolved. Complicating the issue is the fact that parietal cortex contains multiple fields, many of which are involved in the processing of visual spatial information, and some of which appear to be related to the movements of particular body parts. A common approach to studying the role of these fields in transforming sensory information into motor commands has been to consider the frame of reference of the spatial information that each contains. The PI now proposes to study, in addition, whether and how the information is specifically related to arm, eye and head movements (motor specificity). The development of motor specific responses relates to the selection of a particular motor response, a critical stage in the sensory to motor transformation. It is suggested that response selection and reference frame transformation are two distinct and potentially separable components of that transformation. This approach is a significant departure from previous work, which has typically considered only a single motor output. Aim I will characterize the exact relationship between motor specificity and frame of reference for activity in the ventral intraparietal area (VIP). The experiments and their analysis are designed to capture the full range and potential richness of the properties of individual neurons, rather than merely categorizing them as "motor" or "sensory." Aim 2 will examine motor-specific activity in the lateral intraparietal area (LIP) and in the parietal reach region (PRR). We will determine how signals encoding the decision to take one course of action or another interact with signals encoding spatial information. Finally, Aim 3 will challenge an established hypothesis concerning spatial information processing in LIP. The hypothesis has important ramifications for whether processing may be directed towards specific motor outputs. Taken together, the results of these aims are likely to provide new insight into the role of parietal cortex in the sensory to motor transformation in general, and in particular, into the role of specific fields in directing eye, arm and head movements to visual targets.

EIA-0082520
Snyder, Lawrence
University of Washington

ITR: Fluency 1 and Teaching Teachers to Teach IT

Fluency with information technology (FIT) is the ability to effectively use
computers, networks, software and information resources now and in the
future. Fluency is a new standard of IT knowledge formulated by the
National Research Council's Committee on Information Technology Literacy in
their report, Being Fluent With Information Technology. To enable students
to continually learn IT, Fluency provides the foundations (concepts) on
which to build further understanding, and the mechanisms (capabilities) to
understand when more must be learned. The goal of this project is to pave
the way for the widespread offering of information technology fluency
courses in K-12 schools, community colleges and adult education. The
project will allow the creation of a Fluency 1 curriculum whose
presentation stretches over one year, allowing for a paced introduction of
IT topics and time to prepare for and work on IT projects. Fluency 1
contains components of algorithmic thinking and programming and should help
provide teaching guidance that is suitable to assist a present "computer
literacy" teacher in teaching Fluency.

Information Science and Systems (33)
This project addresses the need for access to educational opportunities enabling anyone to become fluent in information technology as defined by the 1999 National Research Council's report "Being Fluent with Information Technology" which defined what Skills, Concepts, and Capabilities everyone should possess for work, citizenship and personal benefit.

This project is creating BeneFIT100, a free, online self-study version for independent learners, based upon the fluency course first developed at the University of Washington in 1999. Developed with high production values, the course includes all the labs and projects of the successful classroom version. To reach students not sophisticated enough for a self-study approach, we also offer an instructor-mediated version of BeneFIT100. We address the need for a Fluency instructor corps by using BeneFIT100 and a specialized workshop to train a nationally selected cohort of future Fluency instructors. Thus, the project creates a new, unique resource of international importance, and demonstrates leveraging it for teacher training.

Description (provided by applicant): Reaching towards a visual target is ubiquitous in daily life. The task seems effortless, yet requires substantial processing to accomplish. Our goal is to better understand the visuospatial information processing underlying action. For this purpose, we use visually-guided reaching in the non-human primate as a model system. Our first aim is to determine the specific contributions of posterior parietal areas to the kinematics and dynamics of visually-guided reaching. We will use a novel method to precisely localize the sites of reversible injections placed throughout the intraparietal sulcus. After each injection we will test animals on a panel of tasks (reaches, saccades and visual search) and then image the site of inactivation. This method is comprehensive and better indicates the true functional contributions of parietal areas than can be achieved through single unit recording. Our second aim is to identify and quantify components of activity in posterior parietal cortex that are related to bimanual coordination. Primates commonly use the two arms together to accomplish tasks that are difficult or impossible to perform with a one arm. Clinical evidence suggests a role of the parietal cortex in bimanual coordination. Our results will help distinguish between two specific models of how bimanual coordination might be manifest at the level of individual neurons. Our third aim is to quantify the activity of posterior parietal neurons during evaluation of targets and decision-making in performing reaches, and to compare that activity to that observed during decision-making for saccadic eye movements. Recent work has suggested a specific model for decision-making for saccadic eye movements. Our results will indicate whether parietal circuits for target evaluation and decision circuits are the same or different for different kinds of action (reaches versus saccades). Achieving these aims will help us understand the early processes involved in sensory to motor transformation, motor coordination, and decision-making. The results will critically inform the devise of rational strategies for aiding recovery from strokes and other central damage, as well as the design of optimal brain interfaces for a new generation of prosthetic devices. PUBLIC HEALTH RELEVANCE The central goal of this proposal is to understand the early processing of visuospatial information for visually- guided reaching. Achieving this goal will help clinicians to understand and ultimately reverse the damage caused by parietal and occipital strokes and other brain trauma. Understanding how the brain generates and represents plans for movement is also critical to the development of promising neuroprosthetics for patients with amputations, spinal cord injuries, and disorders such as amyotrophic lateral sclerosis.

prefrontal lobe /cortex; schizophrenia; brain mapping; short term memory; thalamus; memory disorders; cognition disorders; model design /development; brain electrical activity; pulvinar thalami; disease /disorder model; functional magnetic resonance imaging; behavioral /social science research tag; Macaca; single cell analysis; electrostimulus; neuropsychological tests;

DESCRIPTION (provided by applicant): Connectivity is critical to understanding how the brain works. Currently, most connectivity data is obtained in post-mortem histology studies, which are labor intensive, require the use of many animals, and can only be performed once per animal. Magnetic resonance imaging methods offer much increased efficiency and, because they are performed entirely in vivo, can be repeated as often as desired on the same animal. Diffusion tensor imaging is useful for studying brain connectivity in humans, but in animals, increased resolution and specificity can be obtained by injecting an MR-lucent tracer substance directly into the brain area whose connectivity is in question. Manganese has been used for this purpose in rats and birds, and has been used to trace large subcortical connections in monkeys. We propose to develop a methodology for studying detailed cortical connectivity in living primates, using cerebral manganese injections coupled with MRI analysis and visualization. We will optimize injection techniques, data acquisition and post-processing for intracortical and cortical-subcortical projections. In order to validate our results, we will co-inject manganese with histologic tracers and then compare the results of standard post-mortem histologic analysis with the manganese-enhanced MRI data. Successful development of this method could revolutionize the study of connectivity in monkeys, permitting, for example, precise determinations of how the cortex is rewired in development, skill acquisition and in recovery from lesions. It will also facilitate determinations of areal boundaries in relation to physiologically determined functions, and has the potential to greatly increase our understanding of how circuits in the brain process information. Perhaps the greatest health-related impact of this work will be on rehabilitation after brain injury, since this method will allow us to interrogate brain connectivity as individual subjects respond to and recover from, or fail to recover from, brain insults.

Our long range goal is to improve our understanding of working memory at the level of individual neurons, and to use this information to develop a model of non-human primate model of disrupted working memory that will shed light on the neural basis of the cognitive deficits in schizophrenia. Schizophrenia patients show structural abnormalities in the dorsolateral prefrontal cortex and the thalamus that may be related to their deficits in working memory. We propose to use both established and novel techniques to investigate normal and disrupted working memory function in these and other areas. In particular, we will use whole-brain functional magnetic resonance imaging in the non-human primate to identify neural circuits involved in working memory and to provide a basis for comparison with analogous human data; we will use single neuron recording in the pulvinar nucleus of the thalamus to characterize its role in working memory; and we will use a novel electrical microstimulation paradigm to disrupt function in prefrontal cortex and thalamus during a context-specific working memory task in order to probe the structure of working memory at a single neuron level. By comparing the animal' s performance with that of schizophrenia patients, we will further our understanding of the role that the targeted areas play in mediating the working memory deficits of the patients. Our choice of target areas is guided by the anatomical data of which explores structural deficits in human schizophrenia patients. Our experimental design is and will continue to be informed by findings , which uses behavioral and imaging techniques to study working memory deficits in patients. Once our model is established, we will directly compare patient behavior with that of our model under identical task conditions. In summary, we believe that our strategy of combining neurophysiological, neuroimaging and behavioral techniques with cross-species comparisons will significantly advance our understanding of the neural bases of both working memory and schizophrenia.

DESCRIPTION (provided by applicant): This application addresses Broad Challenge Area (15): Translational Science and Specific Challenge Topic, 15-MH-109: Prefrontal cortex regulation of higher brain function and complex behaviors. The goal of this project is to understand the neural mechanisms that underlie active maintenance in working memory (WM). WM is a core component of most domains of higher cognition and is critically important for executive control. WM dysfunction is thought to play a major role in the cognitive impairments seen in a wide-range of disorders, including ADHD, Alzheimer's disease, and most prominently, schizophrenia. Progress in understanding the neuronal mechanisms underlying normal and pathological WM function would constitute a major advance in basic science, and will serve as a launch pad for studies in clinical populations, such as patients with schizophrenia, for which WM dysfunction is thought to be a major determinant of more widespread cognitive impairment. Our approach is to conduct directly matched imaging experiments in humans and non-human primates using an identical long-duration memory task in both species, and then to conduct multi-unit recording in monkeys again using the same task. Critically, these studies will be designed to provide detailed information regarding the neural mechanisms that result in the decay of stored information over time (i.e., WM delay), and how this translates into behavioral change. Leverage on these issues will be provided by asking which neurons, circuits, and areas have activity that is correlated with the normal loss of stored WM content over time, and/or correlated with the induced perturbations that result from pharmacologic interventions. We will test whether human and non-human primates show similar neuronal patterns in this regard, and we will use multi-unit recording to test specific hypotheses derived from computational models of attractor networks. Our multi- species, multi-method approach will bridge the gap between computational models, single unit recording studies in non-human primates, and human neuroimaging data. By forming this bridge, we will greatly advance our understanding of the mechanisms of working memory in human cognition. Working memory is a high level function that is absolutely critical to normal cognitive function, and is often disturbed in psychiatric illness. We will investigate the normal and pathological function of spatial working memory in the monkey using a variety of methodologies, and then test the relevance of what we have learned in humans using imaging experiments. The result will be a more principled approach to pharmacologic and other therapies for psychiatric illness.

DESCRIPTION (provided by applicant): Resting state networks are a fascinating yet poorly understood new phenomenon in human cognitive neuroscience. Sets of spatially separated regions show correlated slow fluctuations in fMRI BOLD signals even when the subject is at rest. These networks appear to be important in normal brain function: aspects of behavioral performance can be predicted by the ongoing level of slow correlated BOLD fluctuations; brain injuries perturb resting state networks; and multiple clinical disorders, including depression, dyslexia and prosopagnosia, are associated with specific resting state network abnormalities. Currently, resting state data are used to infer functional connections between regions, but little is known about causality, spatial and temporal scale, or the underlying neural substrate of the correlations. A deeper understanding of slow fluctuations and resting state networks has enormous potential for understanding normal and disordered cognition. We seek to better understand the origin and significance of correlated fluctuations by characterizing them at high spatial and temporal frequencies and identifying the electrophysiological signals that are associated with them. The significance of this work is that we will be able to make better use of the fMRI information already being collected, improve diagnosis and perhaps reveal the etiology of several neurological disorders, possibly discover previously unsuspected modes of brain operation, and generally obtain new insight into cognitive processing. Innovation & approach: To obtain these data we propose to use a classical technique, oxygen polarography, in a new way. Guided by resting state fMRI scans, we will insert multiple platinum microelectrodes into a macaque brain in order to verify and characterize correlated fluctuations in oxygen concentration. We will then record simultaneous electrophysiological signals from the same or adjacent electrodes and ask what portion of the electrophysiological spectrum (slow cortical potentials, local field potentials, multi-unit activity) is associated with correlated (resting state network) oxygen fluctuations. This is a new approach to this issue, and we have the required expertise in monkey electrophysiology (L. Snyder, A. Snyder), human fMRI (M. Raichle, A. Snyder), human resting state network analysis (M. Raichle, A. Snyder) and monkey fMRI (L. Snyder, M. Raichle, A. Snyder) to be successful. We have already worked together to establish anatomical and functional monkey fMRI at Washington University, and together we have published data showing resting state networks in the monkey that closely resemble those in humans. PUBLIC HEALTH RELEVANCE: A new methodology, functional connectivity MRI (fcMRI), has recently been applied to diagnose and understand the etiology of a range of diseases and disorders. FcMRI looks at long-distance correlations in brain oxygen to draw inferences about the fundamental structure of the brain and pathological disturbances in that structure. The technique holds great clinical promise, but we currently have very little understanding of why these long-distance correlations exist or what they mean. This grant will provide new information about the origin and interpretation of these correlations, which in turn will greatly increase the amount of clinical information we can extract from the method. In particular, it will improve the diagnosis and understanding of the etiology of the conditions in which it is being currently applied, which include stroke, prosopagnosia and dyslexia.

DESCRIPTION (provided by applicant): The objective of the grant is to provide rigorous and critical training in neuroscience to a diverse cohort of students by providing support for 8 funded positions, or 40 student-years, in the Washington University Neuroscience Program. Specifically, this proposal will support eight early-stage trainees annually for up to three years. Our Program holds long-standing commitments to cutting-edge research, to interdisciplinary education, and to providing modern career development. The Program currently has 90 students, with an average time to the Ph.D. of 5.7 years and an attrition rate of 7.5%. Of the students who have graduated from our Program since 1999, over 70% have remained in academic biomedical research. We seek to be a Program that responds to changes in the research environment by helping our students to pursue important and innovative problems and concepts, to adopt new techniques and to communicate effectively with their peers and the general public. The curriculum and research environments remain broad and deep, combining expertise in molecular, cellular and systems-level approaches to the study of neural function and dysfunction. The Program will continue to recruit and retain talented, diverse students through innovative and dedicated coordination with the University and partner schools. Major new initiatives aimed at accomplishing these goals include: 1) the introduction of a series of grant-writing workshops for students throughout their graduate career, 2) improved training in statistics for students at different levels of preparation, 3) increases in the number of participating Departments and faculty, 4) the introduction of new journal clubs and courses, 5) the introduction of a new second-year Pathway that provides advanced training in human genetics and behavior for neuroscientists, 6) enhanced interactions among students and faculty with regular individual and group advising and new group discussions with the Directors, 7) expanded assessment tools for evaluation of all program elements, 8) new partnerships for trainee recruitment with particular focus on mechanisms to attract and retain a diverse neuroscience community, 9) a novel student-run career development group dedicated to providing intensive yet short-term experiences in biotechnology business, and 10) an organization dedicated to neuroscience outreach and communication of science to the general public. These initiatives will ensure our students remain at the forefront of developments in neuroscience research, teaching and out-reach.

ABSTRACT Oxygen levels within the human brain fluctuate without any apparent external driver. Unexpectedly, these intrinsic fluctuations are correlated among distant regions, forming resting state networks. These networks appear to be relevant to brain function. Resting state data can provide evidence for functional connections between brain regions. Aspects of behavioral performance can be predicted by the ongoing level of slow correlated BOLD fluctuations. Finally, multiple neurological and psychiatric disorders including autism and schizophrenia are associated with abnormalities in resting state networks. Despite their potential importance for understanding normal and disordered cognition, resting state networks remain a poorly understood phenomenon in human cognitive neuroscience. We seek to better understand the origin and significance of correlated oxygen fluctuations by characterizing them at high spatial and temporal resolution and identifying the electrophysiological signals associated with them both at rest and during task performance. We will use oxygen polarography in a novel way. Guided initially by resting state fMRI scans, we will insert multiple platinum microelectrodes into a macaque brain to verify and characterize correlated fluctuations in oxygen concentration. We will record simultaneous electrophysiological signals from these electrodes and ask what portion of the electrophysiological spectrum (slow cortical potentials, local field potentials, multi-unit activity) is associated with task-driven and/or with resting-state correlated oxygen fluctuations. To accomplish this, we will exploit the advantages of polarography over fMRI, including co- localized and simultaneous oxygen and electrical signals, higher spatial and temporal resolution, resistance to movement artifacts, and ease of use in awake behaving animals. Our overall aim is to determine the transfer function mapping electrophysiology signals onto oxygen fluctuations, and whether this transfer function is network-specific, depends on the cortical layer being recorded from, or reflects the ongoing behavioral state of the animal (e.g., task-engaged, sleeping and under anesthesia). The clinical significance of this work is that it will lead to improved use of fMRI information for the diagnosis, prognosis and etiology of brain disorders. The scientific significance, at a high level, is that it will inform our understanding of large-scale brain architecture and cognitive processing.

ABSTRACT Our long-term goal is to understand visuospatial processing in the brain and the neural mechanisms underlying visually-guided movements. Most actions involve coordination across multiple body parts, and so understanding coordination is instrumental to understanding visuomotor processing in general. We will look specifically at the coordination of the two hands (bimanual coordination) and the eye and hand (eye-hand coordination) in the macaque monkey. In this grant cycle, we focus on the parietal reach region (PRR) in the posterior parietal cortex, a key nexus of sensory, motor and attentional processing. In future cycles, we will extend these studies to other brain areas. Our first aim elucidates the role PRR plays (if any) in bimanual coordination. We focus on ipsilateral limb information in PRR and on interhemispheric communication. We will record from neurons during different types of bimanual coordination tasks, all of which require synchronous behavior of the two hands. We will quantify the synchrony of behavior and then determine whether and how PRR activity (single units and LFP) reflects the type of bimanual coordination task and the level of synchrony that is actually achieved. We will also assess information transfer across the two hemispheres and ask if this varies by task and degree of coordination. Our second aim is similar, but adds in the additional dimension of eye movements, spontaneously deployed by animals as they perform these tasks. We will evaluate whether and how PRR reflects different patterns of eye-hand coordination. Our final aim addresses the coding of static eye and hand positions in a bimanual task. In the previous grant cycle, we described an encoding of the distance between gaze and hand in PRR during a unimanual task. Based upon the format of that encoding, we argued that PRR plays a role in reference frame transformations. Discovering whether and how this encoding is maintained in a bimanual task - when there are two eye-hand distances rather than only one to encode - will provide critical information about what computations can or cannot be performed in PRR, and will provide further evidence for, or against, the hypothesis that PRR is involved in bimanual and/or eye-hand coordination. The paradigms we are developing should allow us to identify signals in PRR and elsewhere that are related to motor coordination. Achieving our aim will advance our understanding of the role of PRR in bimanual and eye-hand coordination and also provide invaluable information about how cross-effector and cross-system coordination is achieved in the brain.

DESCRIPTION (provided by applicant): The objective of the grant is to provide rigorous and critical training in neuroscience to a diverse cohort of students by providing support for 8 funded positions, or 40 student-years, in the Washington University Neuroscience Program. Specifically, this proposal will support eight early-stage trainees annually for up to three years. Our Program holds long-standing commitments to cutting-edge research, to interdisciplinary education, and to providing modern career development. The Program currently has 90 students, with an average time to the Ph.D. of 5.7 years and an attrition rate of 7.5%. Of the students who have graduated from our Program since 1999, over 70% have remained in academic biomedical research. We seek to be a Program that responds to changes in the research environment by helping our students to pursue important and innovative problems and concepts, to adopt new techniques and to communicate effectively with their peers and the general public. The curriculum and research environments remain broad and deep, combining expertise in molecular, cellular and systems-level approaches to the study of neural function and dysfunction. The Program will continue to recruit and retain talented, diverse students through innovative and dedicated coordination with the University and partner schools. Major new initiatives aimed at accomplishing these goals include: 1) the introduction of a series of grant-writing workshops for students throughout their graduate career, 2) improved training in statistics for students at different levels of preparation, 3) increases in the number of participating Departments and faculty, 4) the introduction of new journal clubs and courses, 5) the introduction of a new second-year Pathway that provides advanced training in human genetics and behavior for neuroscientists, 6) enhanced interactions among students and faculty with regular individual and group advising and new group discussions with the Directors, 7) expanded assessment tools for evaluation of all program elements, 8) new partnerships for trainee recruitment with particular focus on mechanisms to attract and retain a diverse neuroscience community, 9) a novel student-run career development group dedicated to providing intensive yet short-term experiences in biotechnology business, and 10) an organization dedicated to neuroscience outreach and communication of science to the general public. These initiatives will ensure our students remain at the forefront of developments in neuroscience research, teaching and out-reach.

ABSTRACT Resting state networks are a fascinating yet poorly understood phenomenon. Sets of spatially separated regions show correlated slow fluctuations in fMRI BOLD signals, most obvious when subjects are at rest. These networks appear to have clinical imporantance: brain injuries perturb resting state networks, and multiple clinical disorders, including depression, dyslexia and prosopagnosia, are associated with specific resting state network abnormalities. Resting state data are used to infer functional connections between regions, but little is known about how neuronal activity gives rise to these networks. Furthermore, despite much speculation, little is known about how resting network state might influence task-related neuronal activity. Understanding the reciprocal relationships between resting state networks and neural activity has the potential to revolutionize our understanding of brain function. The reason that a gap in our knowledge exists is primarily due to the fact that it is difficult to characterize resting state networks without being within an active MRI scanner, and difficult to record spikes from neurons in such an environment (and even more difficult to record from multiple cells at once in such an environment). Alternative methods most involve serial recording of resting state networks and neuronal activity, recording of lower frequency electrical signals (LFP), or the use of optical methods in mouse which provide a close but not exact surrogate of neuronal activity (e.g., calcium signals) and access only to the uppermost layers of cortex. We propose to develop an innovative method to address this issue: high density parallel recording and oxygen polarography using carbon fiber microwires widely dispersed across the cortex of an awake behaving non-human primate. We have already demonstrated long-range correlations using oxygen polarography recorded on standard size micro-electrodes, paired with standard unit-recording electrodes. These correlations resemble resting state phenomenon, but in order to capture and relate the dynamics of neural activity to the dynamics of resting state networks, much denser spatial sampling is required.

ABSTRACT Primates, including humans, are expert at coordinating their arms and eyes in skillful behavior. Our goal is to understand the neural circuitry that underlies the combination of bimanual coordination and eye-hand coordination, which we call ?hand-eye-hand? (HEH) coordination. We are particularly interested in the early planning of bimanual movements, and the role-effector specific areas in the posterior parietal cortex play in that planning. We hypothesize that inter-areal and inter-hemispheric communication is necessary for HEH coordination. For example, the parietal reach region (PRR) controls primarily the contralateral arm. One way for one hand to know what the other is doing, so to speak, is for information to be exchanged between PRR in each hemisphere. The most direct pathway for such communication is through the corpus callosum, a major fiber tract connecting the two hemispheres. The relative accessibility of the corpus callosum provides an opportunity for causal tests of the role callosum plays in particular, and of inter-hemispheric communication in general, in HEH coordination. Lidocaine injections can reversibly block conduction through particular portions of the callosum, and behavior and neuronal activity can be compared in behaving animals before, during and after blockade. We predict that HEH coordination will be impaired when particular ?ber tracts within the callosum are blocked, and that there will be neuronal correlates of that impairment within the brain areas responsible for the behavior. Our ?rst Aim is to establish where in the callosum particular axonal tracts cross, and to verify that we can reversibly block conduction through those pathways. Next, for our second Aim, we will test speci?c hypotheses regarding which behaviors will be affected when particular pathways are blocked. We will consider pathways to and from the parietal reach region (PRR) and the lateral intraparietal area (LIP), an analogous area that codes saccade plans. Animals will perform interleaved, natural unimanual and bimanual reaches and saccades. We will then, in our third Aim, examine activity within PRR and LIP to determine how speci?c neuronal circuits are impacted by the transient loss of speci?c callosal pathways. Bimanual HEH coordination is critical for normal human behavior, yet the neuronal circuits involved remain largely unknown. This work will greatly expand our understanding of how parietal cortex achieves complex yet ?exible coordination of body parts. The information will be relevant to coordination in other effector systems, and will help us design the next generation of brain-computer interfacing prosthetics that can leverage natural coordination patterns and coordinate with existing limbs and eye movements. Further, we will learn fundamental facts about the role of the corpus callosum in the brain. Finally, this work will shed light on the general issue of long range communication across brain areas, and how this communication is related to brain function.

ABSTRACT Short-term working memory is critical for all cognition. It is important to fluid intelligence by definition and is disordered in many psychiatric conditions. It is also an ideal model system for studying the link between the dynamics and functions of neural circuits. Short-term storage requires dynamics that are flexible enough to allow continuous incorporation of new information, yet stable enough to retain information for tens of seconds. Much is known about the neuronal substrate of short-term memory. There is a gap, however, in our knowledge of how neuronal resources are efficiently allocated to store multiple items. This gap is particularly striking given that a multi-item memory task (memory span task) is often used to measure fluid intelligence. Neurons in frontal areas are active during a memory period, and individual neurons are tuned to respond to particular memoranda. It is known that individual cells ramp up or down during a memory period. However, we were surprised to discover in preliminary experiments that 80% of individual cells in memory circuits lose their tuning before the end of a 15s memory period. This loss of tuning occurs at similar times across repeated trials; a neuron that loses tuning at 3s in one trial seldom remains tuned for more than 7s in a subsequent trial, and vice versa. This leads to the question of whether cells with common ?drop-out? times are linked together in a subnetwork, similar to the ?slot? organization often posited to support multi-item memory. We formulated a theory about how these subnetworks might be organized to enact a form of efficient resource allocation that balances demand for memory capacity against memory duration. The primary goal of this proposal is to test the validity of this theory, and more generally probe memory circuits for evidence of functional subnetworks, using a unique combination of long-delay multi-item memory tasks, computational modeling and analysis. In Aim 1, we will test a key facet of our theory: how storage of information may interact with the phenomenon of ?drop-out? to either help or hinder short-term memory storage. In Aims 2 and 3, we will test whether cells with similar dropout times also share other properties, as a way of determining whether they are in fact linked together in subnetworks. In parallel, we will develop modeling and optimization tools to ask how such subnetworks might be enacted in neural circuits, while also engaging the higher-level question of whether subnetworks are in fact a sensible solution to the problem of efficient resource allocation in the first place. A key aspect of the proposal is the integration of experimental and computational methods, including formalisms from information and control theories, so as to build tight links between (i) the observed phenomenology; (ii) the mathematical consistency of the theory; and (iii) how (i) and (ii) might be reconciled mechanistically in the dynamics of neural circuits. Together, these Aims have the potential to change the way we think about the neuronal substrate of short-term memory and how neural circuits are structured to best manage their resources.

Understanding overlap in resting state fMRI networks at the single cell level: a cross-species approach Abstract Resting state functional connectivity MRI (rsfcMRI) is a popular tool to investigate the intrinsic functional organization of the brain into large scale networks. Multiple different lines of investigation have pointed to the importance of densely interconnected `hub' regions for cognition and behavior. However, the functional architecture of cellular circuits in these hub regions is unknown. To study the cellular underpinnings of hub regions, we bring together an interdisciplinary research team to bridge across species and across scales. We start by generalizing recent advances in human fcMRI analyses across species to characterize individualized patterns of network overlap in rsfcMRI data from awake macaque monkeys (Aim 1). This allows us to identify regions of interest for recordings in this same animals from a hub region where two (or more) networks spatially overlap, and from two non-hub regions that strongly contribute to only one of the networks respectively. We then ask whether, at a finer cellular scale, there is true neural coupling between both networks in hub regions, or whether networks that appear spatially overlapping at the resolution of rsfcMRI data are in fact spatially interdigitated rather than overlapping at a finer scale (Aim 2). Lastly, we use electrophysiological recordings to determine whether individual neurons in hub regions integrate information from both overlapping networks (i.e. coupling), or whether neurons dynamically switch their network allegiance from one network to another over time (Aim 3). The outcomes of this proposal have important implications for the modeling and interpretation of human rsfcMRI data. This R34 proposal provides the opportunity to establish a new collaboration and validate our methodology across species. These factors are essential for the next stage of our project, a Targeted Brain Circuits Project R01 proposal, in which we will build on this line of investigation by bridging into behavior to study how fundamental principles of the brain circuits in hub regions form the biological basis of mental processes.