Jonathan D. Cohen - US grants

We wish to understand how the nicotinic acetylcholine receptor functions as a ligand-gated ion channel and also to define the interactions within the post-synaptic memberane that are necessary to maintain acetylcholine receptors in high density clusters restricted to regions of the muscle plasma membrane directly underlying the nerve terminal. Torpedo electric tissue will be fractionated to isolate post-synaptic membranes to be used in studies of receptor structure and function. Functional domains within the receptor to be identified include the site of binding of agonists and competitive antagonists as well as the separate binding site for amine non-competitive antagonists. Binding sites will be labeled covalently by appropriate radiolabeled affinity reagents, and the receptor will be degraded to identify the labeled peptides in terms of the known amino acid sequences of the receptor subunits. To determine how the ligand binding sites are oriented within the three-dimensional structure of the receptor, biochemical and immunological labeling procedures will be used to identify receptor domains exposed at the extracellular surface, buried within the lipid bilayer, or exposed on the cytoplasmic surface. These techniques will also be used to identify regions of each receptor subunit in contact with each other or with non-receptor proteins of the post-synamptic membrane. Receptors in the isolated Torpedo post-synaptic membrane have very restricted motility unless the non-receptor peripheral proteins are removed. We will characterize by biochemical and immunological means the structure of the principal non-receptor protein (43kDa protein) and determine whether interactions between this protein and the receptor restrict motility of the receptor.

This proposal represents a maturation of my interest in schizophrenic thought disorder, and my commitment to a multidisciplinary approach to its study. The program I describe here will provide me with fundamental skills in psycholinguistics and cognitive psychology which are directly relevant to research on schizophrenia, and which have not yet been applied in this field. Schizophrenics' use of language has been characterized as lacking in contextual constraint. Psycholinguists distinguish between various loci of contextual effects in language processing, identifying a lexical stage (access to a word's meaning) and postlexical stages. Maher has described a failure to actively suppress contextually irrelevant associations in schizophrenia, and attributes this to a deficit in postlexical processing. I have proposed an alternative, non-competing hypothesis which suggests that disturbances may also arise within lexical processing which account for similar results. In Phase I of this program, I will develop a parallel distributed computer model of language processing. Variables will be sought which can be manipulated in actual psycholinguistic experiments, and which distinguish between the effects of stimulus persistence and postlexical events on the processing of context. In Phase II, I will perform experiments with normal and schizophrenic subjects using chronometric lexical priming techniques, to test predictions which arise from these computer simulations. Positive findings in patients will be correlated with clinical variables, in order to explore the use of these techniques for diagnostic assessment. Work in Phases I and II will be directed to the testing of specific hypotheses concerning schizophrenic language dysfunction. This will not only allow me to explore the functional characteristics of thought disorder, but will also provide me with a training in theoretical and methodological techniques which will be of general value in psychiatric research.

We wish to understand how the nicotinic acetylcholine receptor (AChR) functions as a ligand-gated ion channel and to understand how drugs and toxins interact with the AChR to alter its function. Torpedo electric tissue will be fractionated to. isolate postsynaptic membranes to be used in biochemical studies designed to: 1) identify functional domains (ligand binding sites, ion channel) in terms of the known amino acid sequences of AChR subunits and to identify differences in structure of those domains that distinguish between the binding of agonists and antagonists or between different conformational states of the AChR, and 2) provide a general description of the three dimensional structure of the AChR in terms of the regions of each subunit exposed at extracellular or cytoplasmic surfaces, at subunit interfaces, or in contact with lipid. Radiolabeled affinity labels and structural probes will be covalently incorporated into AChR, and labeled AChR subunits will be isolated and degraded so that sites of labeling will be determined by N-terminal sequence analysis of isolated labeled peptides. The structural studies will provide a definition of particular amino acids and regions contained within binding sites, at subunit interfaces, or at the protein- lipid interface, but they do not of their own assess the importance of the identified amino acids as determinants of ligand binding affinity or their involvement in the mechanism of channel gating. To address these issues a third research goal is to test models of AChR structure derived from the structural studies by analysis of functional properties of mutant AChRs expressed in Xenopus oocytes. The equilibrium binding affinity of agonists and antagonists will be assessed by radioligand binding assays, while AChR function will be assessed by electrophysiological techniques. Point mutations will be introduced to change amino acids predicted to be important for ligand binding affinity or for channel-gating, and additional mutations will be made of amino acids predicted to be important in the propagation of structural changes from the ACh binding site to the ion channel.

9418982 Cohen It is currently possible to acquire images of the brain so as to reveal the regions that are active while people are performing particular cognitive tasks. However, massive amounts of data (many successive images ) are produced from these neuroimaging studies, and it is not clear how to efficiently analyze all of the functional information that the data contain. With this award, a research group comprised of a psychologist, statistician, computer scientist and a radiologist (headed by Dr. Jonathan Cohen) will be developing new ways to analyze large sets of neuroimages. They will improve techniques for image reconstruction, expand available statistical tools for image processing, and implement these improvements on a supercomputer. This work should greatly advance the non-invasive analysis of brain function.

Prefrontal cortex (PFC) is centrally involved in higher cognitive processes such as planning, problem solving and language, and appears to play an important role in major psychiatric illnesses, such as schizophrenia. Despite its role in uniquely human behavior, and illness, will still know relatively little about the function or organization of this brain structure. Neuropsychological and neurophysiological studies have suggested at least two possible principles by which PFC might be organized: representational content and processing function. However, studies of animals and neuropsychiatric patients provide only indirect information about the normal functioning of PFC in humans. Neuroimaging is another method for studying PFC function, that can provide information about the brain activity of normal human subjects as they are performing cognitive tasks. This project used functional magnetic resonance imaging (fMRI) to study PFC, and to test hypotheses derived from computational models of its function and organization in humans. Three sets of experiments are proposed, that will use fMRI to study activation of PFC during performance of cognitive tasks. The first set of experiments will test the idea that PFC is organized by representational content; that is, that different regions support the processing of different types of information. Patterns of activation will be compared in tasks matched for processing function, but involving different types of information (e.g., memory for object vs. spatial and verbal vs. non-verbal information). The second set of experiments will test the idea that PFC is organized by function, by comparing activation in tasks that are matched for information type, but rely on different processing functions (working memory vs. inhibition). Finally, a third set of experiments will examine the role of PFC in distributed cortical circuits involving other brain regions, such as parietal and temporal cortex. An important motivation for these experiments is to test the hypothesis, stemming from computational models of PFC, that the primary function of PFC is the representation and maintenance of context information. Two strong, testable claims follow from this hypothesis: 1) memory and inhibition reflect the operation of a single underlying processing mechanism subserved by PFC; and 2) PFC is organized primarily by representational content. Neuroimaging findings that show activation of different regions of PFC by different types of information would provide support for our hypothesis, while findings showing activation of different regions of PFC by memory and inhibition tasks (i.e., organization by function) would weigh against our hypothesis. Whether our hypothesis is confirmed or refused, however, the results of these studies will provide detailed new information about the function and organization of PFC. A better understanding of PFC promises to provide important insights into the mechanisms underlying uniquely human cognitive abilities, such as problem solving and language, and their disturbance in diseases such as schizophrenia.

This project focuses on the mechanisms involved in cognitive control. By ~control~, we refer to the ability of the cognitive system to flexibly adapt to its behavior to the demands of particular tasks, favoring the processing of task relevant information over other sources of competing information, and mediating task relevant behavior over habitual, or otherwise prepotent responses. The work we propose is a direct extension of our previous efforts to understand and develop explicit models of this function of PFC, and its impairment in schizophrenia. Previously, we have hypothesized that PFC houses a mechanism for representing and maintaining context information. The primary goal of this project is to develop a more detailed theory of PFC function, and its interaction with other cortical systems involved in cognitive control. In particular, we will focus refinements on the mechanisms underlying active maintenance of representations within PFC, the nature of these representations, and the interaction between PFC and hippocampal systems. We will develop and evaluate gated attractor networks as a mechanism for active maintenance of context information. We will evaluate these systems for their ability to a)provide the capacity to maintain context representations in the face of intervening distractor items; b) be implemented using mechanisms that are consistent with data regarding the neuromodulatory functions of dopamine (DA); c)learn to identify and actively store task relevant representation using a reinforcement learning mechanism; d) account for the performance of both normal and schizophrenic subjects in delayed response tasks, and of frontal schizophrenic patients in the Wisconsin Card Sort task. We will also explore the nature of representations within PFC that support its role cognitive control. We will test the following hypotheses: a)Representations within PFC are categorical and combinatorial, and can be used to represent task relevant information at diverse levels of specificity. B)These properties derive directly from the requirement that they be actively maintained. C) The representational scheme in PFC complements that of the posterior neocortex, and can be used to effectively bias processing within the posterior system in accord with task demands. d) Representations with PFC can develop over training in response to task requirements. E) These characteristics can account for normal and impaired performance in neuropsychological tasks classical associated with frontal function, including verbal fluency and object sorting tasks. Finally we will extend our framework in order to account for the role of PFC in more complex cognitive tasks. Specifically, we will develop an integrated model of PFC and hippocampal function, that incorporates an existing model of the latter, and test whether this can account for: a)cognitive control in novel tasks b) the organizational influence of PFC representations over hippocampal storage and episodic memory c) the distinct contributions of PFC and hippocampus to cognitive impairments in neuropsychiatric disorders.

DESCRIPTION (Taken from application abstract): This is an application for a collaborative R01 grant in response to the Human Brain Project PA-96-002. This project will bring together expertise in cognitive neuroscience, functional magnetic resonance imaging (fMRI), statistics, and computer science to develop advanced tools for the processing and analysis of neuroimaging datasets. The primary goal of the project is to improve the temporal sensitivity of fMRI measurements of the neural activity associated with cognitive processes, and to improve the methods by which data are processed and analyzed. The project has three immediate goals: 1) the conduct of empirical studies, that will characterize the dynamics of the fMRI signal in response to experimentally controlled manipulation of the intensity and duration of specific sensory and cognitive processes; 2) the development of quantitative statistical models of these effects that will permit more sophisticated and detailed interpretations of fMRI data than previously possible, significantly improving its temporal sensitivity; 3) the implementation of these computational intensive statistical methods of high performance computing platforms. Success in this work will provide valuable new information about the relationship between the fMRI signal and the neurobiological (and cognitive) processes it is used to measure, as well as new methods for exploiting this information to improve the temporal sensitivity of this technique. More broadly, it will form the basis for a longer term effort to develop tools that have potential for wide applicability, both to other hemodynamically-based neuroimaging modalities (such as MR-based perfusion techniques, and PET), and to other content areas, such as studies of clinical and developmental populations. This project represents a tightly-integrated interdisciplinary effort to advance the state-of-the-art neuroimaging methodology along both scientific and technological dimensions. Success will allow us to better characterize the dynamics as well as the anatomic location of functional activity in the brain. This will significantly improve the sophistication of the questions that can be asked, and productively answered, about the neural bases of cognitive functions and their impairment in neuropsychiatric disorders.

DESCRIPTION: (adapted from Applicant's Abstract) The long range goal of this research is to understand, in molecular terms, how synapses form during embryonic life and how they are maintained in adult animals. These events can be studied with great precision at the vertebrate neuromuscular junction, a synapse whose functional, morphological and biochemical phenotype is understood better than any other chemical synapse. This proposal is focused on ARIA, a polypeptide that was purified based on its acetylcholine receptor inducing activity in embryonic skeletal myotubes. ARIA probably acts at developing and mature junctions to promote the synthesis of ACh receptors and their accumulation in the postsynaptic muscle membrane. One immediate goal of the proposal is to study the processing of ARIA: how the synthesis is regulated in motor neurons, how it is transported to the motor axon terminals and how it is displayed in the synaptic cleft. Membrane binding sites will be studied in detail to determine which type correlates best with physiological responses. Another goal is to characterize ARIA binding proteins in the muscle membrane, and other binding proteins in the extracellular matrix that may modulate its action. Another important set of experiments is designed to define ARIA's spectrum of action at various stages of endplate development. This potent differentiation factor may regulate several genes that encode ion channels in the presynaptic and postsynaptic membrane, proteins in the postsynaptic cytoplasm, and proteins that are destined for the extracellular matrix. It may, in short, be a master switch that regulates synapse specific gene expression. ARIA is synthesized as a transmembrane precursor (proARIA), and several alternatively spliced isoforms have been uncovered. A final goal of this research is, therefore, to determine exactly which isoforms are expressed in motor neurons and the precise role for each one. The planned experiments depend on recombinant DNA, biochemical, electrophysiological and morphological techniques. Identification of ARIA as one of the first trophic factors operating at the neuromuscular junction opens new areas of study regarding the two-way, nutritive relationship between motor neurons and their synaptic partners in the periphery. ARIA was purified from the brain, and it is likely that it exerts powerful trophic effects in the central nervous system as well as in the periphery. Judging from its location in cholinergic neurons throughout the nervous system and in some non-cholinergic cells, our studies at the neuromuscular junction will have immediate import for studies of normal and abnormal development in CNS.

This project proposes the use of the Supercluster II at the Pittsburgh Supercomputing Center to aid in the study of the organization of information in human prefrontal cortex (PFC). This will be done through the processing and analysis of images of cortical activation acquired using functional magnetic resonance imaging (fMRI). There are three computationally intensive primary stages in this processing: the reconstruction of the images from their native Fourier space, the alignment of sets of images, and the statistical analysis of the resulting sets. The parallel architecture of the Supercluster II will be exploited to lower overall analysis time. A master/slave code architecture using the Parallel Virtual Machine library for message passing will be executed in this environment to accomplish these stages of processing. In addition, we are requesting a starter grant on the Cray T3D for this same work.

9871186 With instrumentation provided by the National Science Foundation Drs. Marcia Johnson, Jonathan Cohen and Charles Gross will establish a Center for Cognitive and Behavioral Neuroscience (CCBN) at Princeton University. The Center will focus on understanding the brain mechanisms underlying the cognitive and motivational processes that control behavior, as well as the development of new methods for cognitive neuroscientific research. Research at all levels, from the molecular to the behavioral, will explore the brain mechanisms and subsystems underlying higher cognitive functions (such as attention, memory and decision making), and the closely-related affective processes that govern reward and motivation. Of central interest are the ways in which normal behavior is controlled by higher level goals and regulated by states of arousal, motivation and reward, as well as the ways in which these systems can fail. The Center will couple methods developed during one hundred years of research in physiological and cognitive psychology with important new technologies for investigating the brain. This combination promises to open the way to understand functions of the mind once considered intractable, such as how consciousness and thought emerge from the underlying structures of the brain, what biological systems control emotion and behavior, and what chemical, structural and functional anomalies underlie cognitive deficits and psychopathologies. Several complementary approaches will be employed: fMRI and Event-related Potential (ERP) will be used to examine the structure and function of the brain in conscious human subjects performing cognitive tasks designed to analyze higher mental processes. The separate activity of each of multiple brain cells can be measured. Likewise the expression of particular molecules in individual brain cells as a result of engaging in a specific behavior will also be determined. Two closely related laboratories will be included in the CCBN. An Im aging and Modeling Facility (IMF) will house staff and equipment for the analysis of data sets obtained from fMRI and human electrophysiological studies, and for support of neural network simulation modeling. It will also provide facilities for on site ERP, eye movement and pupilometric recording. During the initial phase of Center development, fMRI data will be gathered at a site less than a mile from the Princeton campus and through collaborations with investigators at other institutions. A closely related Neurotechnology (NTF) Facility will support research at the cellular, molecular and biochemical level. The facility will house modern equipment necessary to conduct analyses of neural substrates of behavior at all of these levels in a variety of mammals, from rodents to primates. Central to proposed research at the IMF is a SGI Symmetric Multiprocessor and associated workstations, microcomputers, networking and software. ERP and eye movement recording equipment, including a Neuro Scan 128 channel SYN-AMPS will also be purchased. A Nikon PCM 2000 confocal microscope system and related computer software will be acquired for use by the NTF. Both graduate and undergraduate students will have access to the laboratories and the instrumentation therefore will serve important educational and training functions.

Project 1 will draw upon empirical and theoretical tools to address a central component of the Center hypothesis: that disturbances of dopamine (DA) in the dorsolateral prefrontal cortex (DLPFC) of patients with schizophrenia contribute in specific ways to the cognitive deficits observed in this illness. Studies will build on previous work establishing that patients with schizophrenia demonstrate a selective deficit in the ability to actively maintain and use context information to appropriately guide behavior. We have hypothesized that this deficit arises from a disturbance of DA activity in DLPFC. Although we have generated behavioral evidence in support of this hypothesis, we have not yet established its relationship to deficits of either DLPFC or DA function. The primary focus of this project will be to establish this relationship, which will provide a direct conceptual bridge between the neurobiological mechanisms involved in schizophrenia and their behavioral manifestations. Under Specific Aim 1 we will conduct functional MRI studies of patients with schizophrenia, to examine the relationship between DLPFC function and performance on cognitive tasks that rely on the processing of context. These studies will be conducted in the same subject populations as those used to study eye movement control and morphometry in Project-Sweeney. Furthermore, they will use the same task that will be used in electrophysiological studies of monkey prefrontal cortex in Project-Olson, thus providing more detailed information about the physiological function of regions engaged by task performance. Under Specific Aim 2, we will examine the influence that pharmacological manipulations of DA function have on behavioral performance, providing information about the role of DA in the processing of context,, and a link to studies of the influence of DA on PFC function to be conducted in Project-Zigmond. Under Specific Aim 3, we will construct computational models of DLPFC and FA function that will be used to interpret the anatomical and physiological data generated about these systems in Projects with regard to their influence on the behavioral functions to be studied in our project and Project-Sweeney. Together, these studies should provide critical new information about the relationship between the neural mechanisms that are impaired in schizophrenia, and their impact on behavior in this illness.

Current fMRI neuroimaging software programs offer the researcher a wealth of analysis methods and tools. However, the incompatibilities in user interface, data format, and computing environment in these tools make it difficult if not impossible for most researchers to take advantage of the full set of tools available for neuroimaging analyses. This software development project will address this problem by creating a common graphical, user friendly environment in which it is easy to incorporate, use and combine analysis tools developed by independent research laboratories. The environment will be a modular, Java-based, CORBA compliant, user extensible client/server architecture with support for parallel processes. Rather than focussing on the development of new analysis tools, we will focus on the development of "software wrappers" which will allow us to substantially improve the interfaces to existing tools without requiring that they themselves be changed. The first specific aim is to write a set of graphical user interfaces (GUIs) to existing fMRI analysis programs that will provide easy to use front-ends for the researcher. We will demonstrate the feasibility, flexibility, and usefulness of this approach by creating wrappers for a significant collection of fMRI analysis software, including locally developed packages such as BRAIN, IFIS, and NIS, which are used by a number of independent laboratories within the Pittsburgh neuroimaging community, and extending to include widely used packages such as AIR and AFNI. To facilitate the integration of these fMRI software packages in the common graphical environment, we will develop two support utilities: 1) format conversion routines for common medical imaging formats such as ANALYZE and DICOM, and 2)synthetic dataset generators to help evaluate analysis programs under a variety of test conditions. The second specific aim is to develop a user friendly graphical "Desktop" that permits the researcher to dynamically link together multiple programs, combine them into one processing stream, and, launch these computations on remote servers from desktop (client) machines. This Desktop will be written in Java and will use CORBA to manage the client/server interactions.

There has been remarkably little research on the mechanisms that detect and signal conflict in cognitive processes, and even less on the role of conflict in the allocation of cognitive control. These questions are central both to our understanding of cognitive control. These questions are central both to our understanding of cognitive function, and to its disturbance in neuropsychiatric disease. The objective of this Center will be to identify and characterize the neural mechanisms involved in conflict and the allocation of control. Several convergent approaches, including computational modeling and formal mathematical analysis, cognitive behavioral testing, functional neuroimaging in adult humans and children, and direct neuronal recordings in non-human species will be used to test the following four Center hypotheses: 1) conflict in processing leads to adjustments in cognitive control; 2) the anterior cingulate cortex is selective responsive to conflict; 3) conflict detection and control are subserved by distinct neural systems; and 4) the influence of conflict on control is modulated by reward. The Center is organized into 7 projects and 3 cores led by highly accomplished scientists in the areas of attention (Casey, Posner, Treisman), memory (Cohen, Johnson) decision making (Kahneman, Shaffir), formal modeling (Holmes, Hopfield), and neurophysiology (Aston-Johnson) decision making (Kahneman, Shaffir), formal modeling (Holmes, Hopfield) and neurophysiology (Aston-Jones, Shizgal) from four institutions (Concordia, Cornell University of Pennsylvania and Princeton). Projects 1-4 will use human behavioral testing and neuroimaging (fMRI). Project 5 and 6 will use behavioral tasks paralleling human studies coupled with neurophysiology and neuroanatomy in monkeys and rats. Project 7 will use computational modeling and dynamical systems techniques to analyze neural network models of the tasks employed in the other projects. Three cores will serve the administrative and technical needs of the Center. These will include an Administrative Core to coordinate Center activities and oversee fiscal and reporting functions, a Neuroimaging Core to coordinate resources for conducting neuroimaging studies of human subjects, and a Computational Core to maintain hardware and software platforms necessary to analyze and visualize large-scale neuroimaging datasets and to conduct computational modeling. In addition to its research mission, the Center will also foster training of researchers at all levels, from undergraduates to participating faculty who are expanding their scientific approaches of cross-project collaborations.

This proposal is for the renewal of a predoctoral and postdoctoral Quantitative Neuroscience Training Program (QNTP) at Princeton University. Neuroscience research is becoming increasingly quantitative. Formal theoretical techniques are essential for understanding how complex, large-scale interactions between neurons give rise to thought and behavior, and advanced quantitative methods of data analysis are necessary for addressing the increasingly large, multidimensional data sets generated by modern brain imaging techniques (e.g., multiunit recording, fMRI). These methods are also necessary for future progress to be made in understanding, diagnosing, treating and, ultimately, curing brain disturbances that give rise to psychiatric disorders. Unfortunately, the mathematical and computational skills required to address these needs are not a focus of standard neuroscience curricula. Princeton's QNTP is designed to address this need, by providing the next generation of neuroscientists with the necessary mathematical and computational skills for measuring, analyzing, and modeling brain function. The establishment of the QNTP sparked several developments at Princeton, that (in turn) have accelerated the pace at which the goals of the QNTP are being met. By bringing Princeton's neuroscientists together with faculty in Physics, Mathematics, Computer Science and Engineering, the QNTP helped to spur the formation of the Princeton Neuroscience Institute (PNI) in 2005. The QNTP also helped to inspire the formation (in 2008) of PNI's free-standing PhD Program in Neuroscience, which strongly emphasizes classroom and laboratory training in basic quantitative and computational methods during its first two years. These developments have made it possible for us to refocus the QNTP from its original purpose (providing a foundation in quantitative neuroscience for trainees who are starting out in this area) to providing advanced training in quantitative neuroscience. Specifically, we will take the most quantitatively-focused subset of our predoctoral and postdoctoral trainees and provide them with the additional tools and training that they need to excel in computational neuroscience research. This training will be accomplished via advanced quantitative and computational neuroscience elective courses that were developed for the QNTP and are taught by leaders in the field, as well as participation in research seminars, journal clubs, and career development activities that are designed to deepen the trainees' knowledge and bolster community among the trainees. PNI faculty have made seminal contributions to quantitative neuroscience, ranging from information- theoretic analyses of neuronal spiking and dynamical systems analysis of decision-making to multivariate decoding of human neuroimaging data. The QNTP has been specifically formulated to bring predoctoral and postdoctoral trainees into contact with this expertise and, through this, to catalyze their transformation into full- fledged computational neuroscientists. As with the prior funding period, we are requesting support for four predoctoral trainees and four postdoctoral trainees (with 2 year appointments for each trainee).

DESCRIPTION (provided by applicant): A fundamental challenge in cognitive neuroscience involves the identification of the mechanisms that allow the flexible, goal-directed behavior that characterizes human performance. Our previous work has led to a theory regarding the role of prefrontal cortex (PFC) in guiding behavior in accord with internally represented goals and intentions. Central to this guided activation theory of PFC is the hypothesis that PFC actively maintains goal representations that guide activation in posterior structures responsible for executing task-relevant behavior. This project has two specific aims. First, we will attempt to further our understanding of the rich, interactive dynamics between PFC representations and activity in posterior cortex. In particular, we will try to find support for our claim that PFC representations modulate task processing in posterior cortex in a continuous and direct fashion. Second, we will examine the mechanisms involved in the flexible updating of PFC goal representations in response to cues in the environment. We propose that the medial temporal lobe may play a crucial role in the forming of episodic associations between environmental cues and task goals that are used to establish effective goal representations in PFC. We will study these issues within the context of the task-switching paradigm, an experimental framework that is ideally suited to our purposes because it entails the repeated establishment and shifting of goal representations needed to guide behavioral responses to ambiguous stimuli. We will develop computational models with the aims of (i) demonstrating the power of the guided activation theory of PFC in explaining a wide variety of phenomena reported in the task-switching literature; and of (ii) generating detailed empirical predictions regarding the neurobiological components underlying the functions of interest here. In parallel, we will conduct a series of fMRI and ERP studies to test these predictions, often contrasting them with predictions made by competing theories. Successful execution of this project will result in a much more detailed understanding of the neural mechanisms underlying cognitive control. Importantly, our models may serve as a much desired foundation for more rational approaches to research on the disturbance of these mechanisms in neuropsychiatric disorders, such as schizophrenia, depression, and addiction, all of which are known to be associated with disruptions of cognitive control.

Drs. Cohen, Darley, and Greene's research funded by NSF on neuroscientific moral psychology was inspired by a puzzling set of moral dilemmas posed by philosophers. Consider the following case: A runaway trolley is headed for five people who will be killed if it proceeds on its present course. The only way to save them is to flip a switch that will turn the trolley onto an alternate set of tracks where it will kill one person instead of five. Ought you to turn the trolley in order to save five people at the expense of one? Most people say yes. Now consider a similar dilemma: As before, a trolley threatens to kill five people. You are standing next to a large stranger on a footbridge spanning the tracks, in between the oncoming trolley and the five people. This time, the only way to save the five people is to push this stranger off the bridge and onto the tracks below. He will die if you do this, but his body will stop the trolley from reaching the others. Ought you to save the five others by pushing this stranger to his death? Most people say no. For over twenty years, moral philosophers have been puzzling over cases such as these, wondering what makes it acceptable to sacrifice lives in some cases but not others. In their research, Drs. Cohen, Darley, and Greene attack these problems from the point of view of social psychology and cognitive neuroscience: What goes on in people's brains that makes them say "yes" to the first case and "no" to the second case? Existing theories of moral psychology suggest strikingly different answers to this question. According to the rationalist tradition in moral psychology, moral judgments are caused by episodes of reasoning and reflection. More specifically, a rationalist would say that people arrive at different answers in these two cases by applying abstract moral principles that explain why these cases are importantly different. A more recent trend in moral psychology places increased emphasis on emotion. According to an emotivist model, differences in emotional response are to explain people's divergent answers in these two cases. Drs. Cohen, Darley, and Greene believe that rationalists and emotivists are both partly correct. Their NSF-supported research is aimed at understanding how emotional and "cognitive" processes interact to produce moral judgments. Their research uses both traditional methods such as questionnaires and measurements of reaction time in conjunction with cutting-edge neuroimaging techniques that allow them to see what is going on in people's brains while they make moral decisions.

This research has natural connections to matters of both private and public concern. First, understanding the psychological and biological bases of human morality is of fundamental humanistic importance. Like research concerning the origins of life on Earth or the large-scale structure of the universe, this research addresses questions that are of intrinsic interest to people around the world. Our capacity for moral judgment is central to our humanity, and yet it is not well understood by science at this time. This research is an important step toward remedying this ignorance. Second, moral judgment is of immense practical importance. Many of the great public debates of our time such as those concerning abortion, stem cell research, the limits of justifiable war, the appropriate response to terrorism, etc. exist because different people have different intuitions about these and other matters of right and wrong. To make progress on these issues it may be useful, if not essential, to understand the psychology and underlying biology that produces moral judgments, and different moral judgments in different people. One of the goals of this research is to study culturally-based differences in moral judgment, which has the additional benefit of ensuring the participation of groups who are underrepresented in American science. This research will also explore differences in moral judgment based on gender and individual temperament. At the same time, however, this research is aimed at understanding that which is universal in human moral judgment.

DESCRIPTION (provided by applicant): This is a competing renewal of an R01 focused on understanding the neural basis of cognitive dysfunction in schizophrenia. In the previous funding period we tested the hypothesis that schizophrenia patients had a deficit in the representation and maintenance of context, which could account for a range of cognitive disturbances in schizophrenia, and was associated with a disturbance in the function of the prefrontal cortex. Our results have been strongly supportive of these hypotheses, and in addition suggest that the prefrontal deficits in schizophrenia may be functionally and anatomically specific. Specifically, it appears that while context processing functions of the dorsolateral prefrontal cortex (DDLPFC) are impaired other functions within working memory, such as phonological rehearsal processes associated with the function of posterior ventrolateral prefrontal cortex (VLPFC) are intact. The goal of the renewal is to extend our understanding of the functional anatomy of impaired cognition in schizophrenia by pursing 2 aims. The first Aim is to use event-related fMRI and 3 cognitive tasks to test the hypothesis that impaired cognitive control in schizophrenia primarily involves functional disturbances of the DLPFC and anterior cingulate cortex, 2 highly interconnected regions of the frontal lobe which appear to contribute complementary functions to the regulation of cognitive control, while more superior and inferior frontal regions are functionally intact. The second Aim is to use multiple methods from cognitive neuroscience, including computational modeling, event-related fMRI and high density ERP, to better characterize the functional contribution of one (1) of the above regions, the ACC, to impaired cognitive control in schizophrenia. The studies will be completed in never medicated first episode schizophrenia patients, to address potential confounds of medications and chronicity. Successful completion of this work will increase our understanding of the functional anatomy and mechanisms of cognitive dysfunction in schizophrenia that should help guide the development of treatments for this disabling and relatively treatment refractory aspect of the illness.

[unreadable] DESCRIPTION (provided by applicant): This project will study the optimization of speeded decisions, and the control and monitoring mechanisms that serve this optimization by balancing performance costs and benefits. These lines of work will be pursued under two specific aims. First, we will build on theoretical work from the previous period, which used the drift diffusion model and neural network modeling to produce detailed predictions about behavior and the dynamics of neural mechanisms underlying control. Here, we aim to test these predictions using behavioral, fMRI, and electrophysiological techniques, in coordination with continued modeling and analysis work in Project 6. These studies will address two-alternative forced choice tasks in which speeded decisions must be made between a correct (rewarded) and incorrect (unrewarded) alternative, tasks that will also be explored within human developmental and non-human primate populations in Projects 2, 4, and 5. In a second, complementary line of work, we will extend our investigation to decisions that require a more graded evaluation of the relative costs and benefits of alternative courses of action. Many decisions do not involve a simple choice between correct vs. incorrect (rewarded vs. unrewarded) alternatives, but rather require an evaluation of the relative costs and benefits of the options. We will use behavioral, imaging, electrophysiological, and pupillometric methods to characterize the neural mechanisms involved in evaluating costs and benefits in such circumstances, and how these evaluations are combined to compute utility and control decision-making within and between tasks. Experiments in this second line of studies will parallel developmental studies in Project 4 and neurophysiological studies in Project 5 examining the decision between a well-defined alternative versus the opportunity to explore other alternatives. [unreadable] [unreadable]

The Computational Core will serve the Conte Center in five primary ways: 1) It will provide access to a recently upgraded computing facility, which includes a compute cluster consisting of 64 dual-processor nodes, a dedicated file server with 9 IB of storage, a tape backup system, and a high speed network. All projects involve either computationally intensive analyses of large datasets or mathematical model simulations, so they will benefit directly from the powerful, new computing resources available;2) It will also provide access to a wide range of generalpurpose software (used by all projects), dedicated neuroimaging data analysis packages (projects 1,2,3 and 4) and mathematical simulation environments (projects 3 and 6). Customized software development services will be available for specific project needs as well;3) The Computational Core will provide a flexible, easy to use, and secure platform for sharing data, which will facilitate the exchange of datasets between projects. All projects, having investigators located at four different sites, will benefit from a robust data sharing initiative;4) It will provide new data analysis methods aimed at extracting and exploiting the information contained in spatially distributed patterns of brain activity. These methods will be used to analyze data collected in projects 1, 2 and 3, and 5) It will also provide training and support for all Conte Center participants on how to use the computing facility, its software, data sharing tools, and new data analysis algorithms. The training, as in previous years, will come in the form of an annual tutorial operated in conjunction with the Administrative and Neuroimaging Cores, as well as through day-to-day support. Overseeing these aims will be a Director (Singer), who has extensive experience in software design and computer hardware. Assisting him will be a Co-director (Takerkart), a signal-processing engineer who has acquired an in-depth knowledge of neuroimaging methods during his five years at the CSBMB, and a Systems Manager (Tengi) who has efficiently fulfilled this role for the last six years. The Center PI (Cohen) will work closely with the Director to facilitate coordination of core activities with those of the other cores, the projects, and in facilitating new collaborations.

This project will examine the development of cognitive and neural processes underlying reward-based decision-making. Incentive-based rewards can bias decisions (Montague et al.,1996) with top-down goal directed processes (cognitive control;Miller &Cohen, 2001) or by bottom-up processes related to the predictability of information (reinforcement learning;Montague et al., 1996;Reynolds et al., 2001). We will use the differential development of these systems over childhood to further dissociate the cognitive and neural processes underlying reward-based decisions. We will use a simple two-choice decision task and manipulate reward magnitude, reward uncertainty and reward rate in making a particular response choice. Formal models of reinforcement learning and principles of decision making (diffusion model) together with functional neuroimaging will be used to precisely characterize cognitive and neural processes underlying decision making and constrain a prior hypotheses and interpretations of results. We hypothesize that decisions in childhood are largely driven by bottom-up (reinforcement mechanisms) rather than top-down goal directed behavior due to an inefficient, less mature cognitive control system in children. As such decisions will be optimized by immediate reward (exploitive), but uncertainty or delay in reward will lead to suboptimal performance. Specific Aim 1) To examine the development of cognitive and neural processes underlying decisions as a function of reward magnitude. Specific Aim 2) To examine the development of cognitive and neural processes underlying decisions as a function of reward frequency and uncertainty. Specific Aim 3) To examine the development of cognitive and neural processes underlying decisions as a function of reward rate.

This project will study the optimization of speeded decisions, and the control and monitoring mechanisms that serve this optimization by balancing performance costs and benefits. These lines of work will be pursued under two specific aims. First, we will build on theoretical work from the previous period, which used the drift diffusion model and neural network modeling to produce detailed predictions about behavior and the dynamics of neural mechanisms underlying control. Here, we aim to test these predictions using behavioral, fMRI, and electrophysiological techniques, in coordination with continued modeling and analysis work in Project 6. These studies will address two-alternative forced choice tasks in which speeded decisions must be made between a correct (rewarded) and incorrect (unrewarded) alternative, tasks that will also be explored within human developmental and non-human primate populations in Projects 2, 4, and 5. In a second, complementary line of work, we will extend our investigation to decisions that require a more graded evaluation of the relative costs and benefits of alternative courses of action. Many decisions do not involve a simple choice between correct vs. incorrect (rewarded vs. unrewarded) alternatives, but rather require an evaluation of the relative costs and benefits of the options. We will use behavioral, imaging, electrophysiological, and pupillometric methods to characterize the neural mechanisms involved in evaluating costs and benefits in such circumstances, and how these evaluations are combined to compute utility and control decision making within and between tasks. Experiments in this second line of studies will parallel developmental studies in Project 4 and neurophysiological studies in Project 5 examining the decision between a well-defined alternative versus the opportunity to explore other alternatives.

DESCRIPTION (provided by applicant): The work described in this proposal seeks to further our understanding of the neural mechanisms underlying impulsivity, and their modulation by social influences. Impulsivity lies at the heart of a large number of maladaptive behaviors, ranging from ones common in every day life (e.g., bad eating habits, and failures to adequately save) to ones closely associated with clinical disorders such as drug addiction, ADHD and aggressive behavior. The propensity for impulsivity varies substantially over the life cycle, as does its susceptibility to social influences (such as authority figures, peer-pressure, and advertising). We will study impulsivity in the context of intertemporal decision making. Intertemporal decisions involve choices between two rewards available at different times. It is well recognized that people discount rewards more steeply over the near term than over longer terms. That is, rewards that are available immediately have disproportionately high value to us, and delaying them (e.g., by a week) devalues rewards substantially more than when a similar delay is imposed on future rewards (e.g., from one week to two weeks). One explanation for this behavior is that human discounting involves two separate systems, one that heavily devalues the future, and another that is more sensitive to future reward. In previous work, we have found evidence that distinguishable neural mechanisms may be associated with each of these systems, and that their relative activity correlates with the choices that people make between immediate and future rewards. In this proposal, we will conduct studies to refine our understanding of these neural mechanisms, their relationship to intertemporal choice, to individual differences in impulsivity, and their modulation by social influences. Under our first two aims, we seek to manipulate each system individually, in an effort to isolate the contributions that each makes to intertemporal choice. Under a third aim, we will examine the influence that social factors have on the functioning of these mechanisms, in an effort to better understand the mechanisms that mediate social influences on impulsive behavior. An understanding of the neural mechanisms underlying impulsivity, and how these are impacted by social influences, promises to provide insights that will be useful in designing interventions, both at the individual and social levels that better mitigate the costs of impulsivity to the individual and society.

[unreadable] DESCRIPTION (provided by applicant): The mission of Princeton's Center for the Study of Brain, Mind and Behavior (CSBMB) is to support multidisciplinary research on the neural mechanisms underlying mental function. Center researchers come from a variety of disciplines, including mathematics, physics, chemistry, engineering, neuroscience and psychology. A primary focus of their work has been the use of fMRI to study the pattern and dynamics of brain activity associated with mental functions such as visual perception, attention, working memory, long term episodic memory, and decision making. CSBMB investigators have been at the forefront of this field, including the development of new methods for analyzing fMRI data. For example, they have led the development of multivariate methods of pattern analysis, and have demonstrated their ability to distinguish between cognitive states not possible using standard methods of analysis. These new methods require larger datasets, and substantial expansion of the raw data, placing increasingly heavy demands on computational resources. At the same time, growing interest in these efforts has attracted additional investigators to the CSBMB and its facilities. This has been further accelerated by the recent formation of Princeton's Neuroscience Institute and its plans for faculty expansion. These developments are placing strains on the CSBMB's computing facilities, the current limits of which are beginning to constrain the research efforts of CSBMB investigators. In this proposal, we request support to expand this facility by substantially increasing the storage capacity of its file server, and upgrading bandwidth on network connections used to access this server. The CSBMB and the University have committed to match these requests with an expansion of the memory for the CSBMB compute server, and to continue to cover all expenses for the maintenance and support of this computing facility. [unreadable] [unreadable] Work supported by the CSBMB aims to deepen our understanding of the neural mechanisms underlying mental functions that are disturbed in a wide range of clinical conditions, including depression, anxiety disorders, schizophrenia, neurological impairments, aging, and drug addiction. CSBMB investigators are supported, in this work, by grants from a number of NIH institutes, including NIMH, NINDS, NIDA, NIA and NIE. Progress in this work promises to lead to more sophisticated and more effective approaches to the diagnosis, treatment and ultimately cure of mental disease, in just the way that our understanding of basic mechanisms of cardiovascular function have led to more sophisticated and effective treatment of heart disease. The current proposal will support the research efforts of CSBMB investigators by providing them with the technological infrastructure necessary to pursue more sophisticated neuroimaging studies, to conduct more detailed analyses of the data they generate, and to construct more realistic models of neural function used to interpret these findings. [unreadable] [unreadable] [unreadable]

This Major Research Instrumentation award permits Dr. Jonathan Cohen and four co-investigators to purchase a high-performance computing instrumentation (3,584 cores; 2TB/core; 100TB flash storage) to be used by faculty, postdocs, graduate students and undergraduates within the Princeton Neuroscience Institute (PNI). The instrumentation will allow the analysis of human brain imaging data at a speed and scale not previously possible.

The collaborating researchers are cognitive neuroscientists and computer scientists at Princeton with complementary expertise in human brain imaging and large scale computing. Two primary research objectives are proposed, building on recent progress in applying multivariate pattern analysis (MVPA) methods from machine learning to detect neural signals that correspond to internal mental states, such as perceptions, memories and intentions that are otherwise not accessible to direct observation. To date, use of MVPA has been restricted to the "offline" analyses" after data have been fully collected. However, a growing and powerful use of brain imaging is to give participants feedback about their brain states in real time, allowing them to use this information to better control brain function (e.g., providing feedback about pain areas as a way of learning to control chronic pain). Such real-time feedback methods could be greatly enhanced by adding MVPA. However, this has been computationally intractable until now. Objective 1 addresses this challenge, by inserting a high performance computing system into the brain scanning pipeline. This will be tested in an experiment that uses MVPA to detect patterns of brain activity associated with sustained attention, allowing us to provide real-time brain-based feedback to improve attentional abilities (with potential educational and health benefits).

Objective 2 focuses on another major advance in brain imaging, in which correlations between areas of activity are analyzed, rather than areas of activity in isolation of one another. Such correlations - often referred to as "functional connectivity" - are likely to reveal more about how the brain actually functions, by providing critical information about the interactions between areas. At present, virtually all approaches to functional connectivity focus on the correlations among a limited set of brain areas of interest. However, a more powerful approach would be to examine the correlation of every area with all others. This requires computing the whole-brain correlation matrix. The analysis of such high dimensional data would be further enhanced by applying MVPA to patterns of correlation. However, doing this further increases computational demands. Applying this approach to a routine brain imaging dataset, using currently available instrumentation, would take 880 years to complete. The work under Objective 2 addresses this challenge, by coupling massively parallel computing with sophisticated software optimizations. Doing so can bring previously intractable problems into the range of practicality. These methods will be tested in an experiment that seeks to identify neural representations of intentions, and their influence on brain mechanisms responsible for executing these intentions.

DESCRIPTION (provided by applicant): This proposal requests a new high-performance computing (HPC) system for the Princeton Neuroscience Institute (PNI). This is needed to meet both the growing user base and the rapid growth in data analysis and storage demands of leading edge neuroscientific research. The PNI's mission is to support multidisciplinary research on the neural mechanisms underlying mental functions such as perception, attention, memory, learning, decision making, and cognitive control. PNI places a strong emphasis on the development of formally rigorous theory, and quantitatively sophisticated approaches to the analysis of neuroscientific data. Their work makes use of state-of-the-art methods for recording neural activity at all levels of analysis, from single- and multi-unit recordings of neurons in non human species to whole brain fMRI and EEG studies in humans. The size of the data sets generated by these methods has grown explosively over the last decade, and the methods needed to analyze them have become increasingly computationally demanding. At the same time, the PNI user base has grown and will continue to grow substantially over the next several years, from 15 at its inception in 2005, to 21 at present, and to an expected 26 as it occupies its new building presently under construction (and due for occupancy in 2013). These factors have conspired to place severe strains on existing PNI computing facilities, the limits of which are now constraining the research efforts of its investigators. To meet these needs, this proposal requests support for a new HPC system comprised of a 52 node cluster (with 832 cores) and a 540 TB storage system. This new system will be housed in Princeton University's newly constructed High Performance Computing Research Center (HPCRC), co-administered with the Princeton Institute for Computational Science and Engineering (PICSciE), and linked to other powerful computing systems at the HPCRC by way of a high performance GPFS storage grid. This will allow PNI investigators to leverage the availability of considerable additional CPU power at the HPCRC with transparent access to their data, as well as the expertise of PICSciE staff in parallel computing. Work supported by the PNI aims to deepen our understanding of the neural mechanisms underlying mental functions that are disturbed in a wide range of clinical conditions, including depression, anxiety disorders, schizophrenia, neurological impairments, aging, and drug addiction. PNI investigators are supported, in this work, by grants from a number of NIH institutes, including NIMH, NINDS, NIDA, NIA and NEI. Progress in this work promises to lead to more sophisticated and more effective approaches to the diagnosis, treatment and ultimately cure of mental disease. The current proposal will support the research efforts of PNI investigators by providing them with the technological infrastructure necessary to pursue more sophisticated neuroscientific studies, to conduct more detailed analyses of the data they generate, and to construct more realistic models of neural function used to interpret these findings.

PROJECT ABSTRACT This proposal is to support undergraduate participants in the Princeton Neuroscience Institute Summer Internship Program. The PNI summer undergraduate research program is geared towards undergraduates who have a strong passion for scientific research and are seriously considering graduate studies in neuroscience. This summer internship provides both education and hands-on research experience in the field of neuroscience. Participants closely collaborate with students and faculty at PNI on original research projects, and thereby gain invaluable first-hand experience as a practicing neuroscientist in a top tier laboratory. PNI summer interns also participate in collective educational activities, including weekly foundational lectures created by PNI faculty specifically for the interns, as well as lab meetings, journal club reading groups, and special forums on topics such as career advice and graduate school application preparation. Princeton has welcomed summer undergraduate students since 2013, and with this proposal, we intend to build upon the successes of the past to establish a comprehensive program to prepare undergraduate students for a career in neuroscience. We will identify promising students, from other institutions, during their sophomore and junior year. We will then provide a structure for these students to have a total experience to make them competitive for a research career in neuroscience.

PROJECT SUMMARY The skin serves a dual function as an immunological barrier and a sensory interface between the body and environment. Protection against invading pathogens is accomplished by coordinated interactions between immune cells in the skin whose aberrant activity can provoke pathologic inflammation. Increasing evidence has demonstrated a unique role of pain sensing fibers, or nociceptors, in cutaneous immune responses. Briefly, Trpv1+ nociceptors were found to be required for IL-23-dependant production of IL-17 by dermal gamma delta and CD4 T-cells. Mice deficient in these nociceptors are more susceptible to C. albicans skin infection but have diminished pathology in an IMQ mouse model of Type-17 psoriasaform inflammation. The requirement of nociceptors for cutaneous immunity suggests that neuronal activation by itself may be sufficient to affect the balance between protective host defense and pathologic inflammation. To test the sufficiency of nociceptors for skin inflammation, we developed an optogenetic murine model which allows for selective activation of nociceptors with high temporal and spatial precision. We found that activation of Trpv1+ nerves is sufficient for skin inflammation involving production of IL-23, IL-6, and TNF? and infiltration of IL-17 producing T-cells. Interestingly, activation of distinct nociceptors expressing MrgprD induces expression of Type-2 but not Type-17 cytokines. Taken together, the sufficiency of nociceptors to modulate cutaneous immunity highlights nociceptors as a therapeutic target to enhance host defense or limit pathologic inflammation in the skin. We hypothesize that activation of Trpv1+ nociceptors is sufficient to activate a cascade of events starting with CGRP activation of cDC2 which in turn release cytokines that induce IL-17 production from T-cells. We anticipate that this Type-17 immune response will provide protection to epicutaneous S. aureus infection. Finally, we hypothesize that induction of Type-17 immunity is specific for Trpv1+ nociceptors whereas activation of MrgprD-expressing fibers will be sufficient for Type-2 immune responses. We will address these questions in three specific aims. Aim 1: Test the hypothesis that Trpv1+ nociceptor activation drives IL-23, IL-6 and TNF? secretion from cDC2. Trpv1-Ai32(Chr2-YFP) mice will be cutaneously photostimulated. Tissue will be analyzed by flow cytometry and qPCR for IL-23, IL-6 and TNF?. Finally, a series bone marrow chimeras will be generated allowing for the depletion of cDC2 to demonstrate functional non-redundancy. Aim 2: Test the hypothesis that Trpv1+ nociceptors are sufficient and required for host-protection against S. aureus infection. Mice with ablation of Trpv1+ nociceptors or photostimulated Trpv1-Ai32 mice will be epicutaneously infected with S. aureus. CFU and immune parameters for Type-17 inflammation will be analyzed by flow cytometry. Aim 3: Test the hypothesis that MrgprD+ nociceptors are sufficient for Type-2 inflammation. MrgprD-Ai32 mice will be optogenetically activated and assayed for Type-2 inflammation by flow cytometry and mRNA expression.

Every day people have to choose between getting something immediately or getting something even better in the future. Saving for a vacation, retirement planning, and even working towards completing a large project all require foregoing immediate rewards to achieve long-term goals. Psychologists, neuroscientists, and economists have studied how people make these decisions, exploring how the weight put on the present and the future vary across individuals and their situations. Crises, such as the spread of COVID-19 in the United States, involve a unique configuration of stresses that may influence the way that people think about the present and the future. In such crises, short-term thinking can have detrimental consequences.

The research team is exploiting a unique and urgent opportunity to document how people?s choice between immediate and delayed rewards changes during a crisis. In January 2020 the team collected a large dataset on inter-temporal choice from over 3,000 participants. The researchers are using this dataset as a baseline for examining how people shift between long-term and short-term during the COVID-19 crisis. Heterogeneous infection rates and remediation strategies in different regions provide an unprecedented natural experiment for examining the impact of these factors on how people make decisions. By collecting an equivalent dataset at multiple points in the progress of the crisis, together with information about local conditions and stress levels, the reseaach team can can explore how these factors influence people?s decisions.

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.

Project Summary Paralleling the growth of neuroscience research, there has been an explosion in the development of computationally explicit models of the functions of core brain subsystems. Unfortunately, however, there has not been a commensurate development of the tools needed to share, validate, and compare such models, or integrate them into models of system-level function. Such sharing, evaluation, and integration are necessary if computational modeling efforts are to be useful not only in generating reliable and accurate accounts of how brain subsystems operate, but also of how they interact to give rise to higher cognitive functions, and how disruptions of such interactions may give rise to disturbances of mental function observed in psychiatric and neurological disorders. This proposal seeks to meet this need by developing PsyNeuLink: an open source, Python-based software environment that makes it easy to create new models, import and/or re-implement existing ones, integrate these within a single software environment that will facilitate head-to-head comparison of comparable models, the assembly of complementary models into system-level models, and serve as a common repository for the documentation and dissemination of such models for both research and didactic purposes (i.e., publication, education, etc.). These goals will be pursued under two Specific Aims: 1) Extend the scope of modeling efforts that PsyNeuLink can accommodate by: i) enhancing its application programmer interface (API) used to add new components and interfaces to statistical analysis tools and other modeling environments (such as PyTorch, Emergent and ACT-R; ii) enriching its Library by adding PsyNeuLink implementations of influential models of neural subsystems; and iii) developing a publicly available workbook of simulation exercises as both an introduction to PsyNeuLink and for use in Cognitive Neuroscience and Computational Psychiatry curricula. 2) Accelerate PsyNeuLink by developing a custom compiler that preserves its simplicity and flexibility, while dramatically increasing its speed, to make it suitable for simulation of large and complex system-level models, and for parameter estimation, model fitting, and model comparison. This project will exploit the power and accelerating use of Python, and modern just-in-time compilation methods to develop a tool designed specifically for the needs of systems-level Cognitive Neuroscience and Computational Psychiatry. This promises to open up new opportunities for research at the systems-level ? a level of analysis that is crucial both for understanding how human mental function emerges from the interplay among neural subsystems, and how disturbances of individual neural subsystems impact this interplay, disruptions of which are almost certainly a critical factor in neurologic and psychiatric disorders.

The NSF Convergence Accelerator supports use-inspired, team-based, multidisciplinary efforts that address challenges of national importance and will produce deliverables of value to society in the near future. Accelerating convergence in science and technology depends on the ability to represent and share not only data, but also theories and models in the most objective, transparent, and reproducible way possible. This project will develop a Model Description Format (MDF) that can be used for computational models that span from neuroscience and psychology to machine learning, and that can serve as the foundation for extensions that serve an even broader scope of models in population biology and the social sciences.

Such an MDF would have numerous benefits, both scientific and technological, including: dissemination and validation of model reproducibility; migration of models across domains (e.g., use of models of brain function in machine learning applications); integration of models at different levels of analysis (e.g., biophysically-realistic neural models into models of cognitive function, cognitive models as agents in population level models); exploitation of complementary strengths of existing packages (e.g., design in a familiar environment but execute in one with better tools for parameter tuning and/or data-fitting); and more efficient development of new tools, by providing developers with a representative diversity of models, all in a common format.

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.