Course of Study
During the first year of their Ph.D., all students take the Neuroscience Core Course. The goal of this course is to provide a common foundation, so that all students have a strong knowledge base and a common language across the breadth of Neuroscience, a highly diverse and multidisciplinary field. To the extent possible, the course aims to teach an overview of all topics through a mix of hands-on laboratory experience, lecture, and computational modeling.
Ph.D. students must also
- Choose and take one elective course from those listed below.
- Rotate, during the first year, in up to three laboratories, participating in research projects during each rotation.
- Pass their general exam, which will include both a breadth component and a thesis proposal depth component, by the end of their second year.
- Most importantly, students must carry out original research leading to a Ph.D. thesis.
Across the board, from molecular biology to physics to psychology, Princeton's world-class faculty is particularly strong in quantitative and theoretical investigations. The same is true in Neuroscience. In recognition of this, a Quantitative and Computational Neuroscience track exists within the Neuroscience Ph.D. Students in this track must fulfill all the requirements of the Neuroscience Ph.D. In addition, their electives should be in quantitative courses, and their Ph.D. research should be in quantitative and/or computational neuroscience. The QCN track is supported by the T32 training grant in Quantitative Neuroscience from the NIMH.
Neuroscience Core Course
The year-long Core Course is the foundation for coursework in the Neuroscience Ph.D. In terms of time and effort, this course counts as two regular courses for each of the two semesters. Lectures, laboratory work, and computational studies are intertwined throughout the course. We expect different students to come in with very different backgrounds; as a result, no previous experimental or computational experience with any particular technique is assumed.
The course content was designed based on the three training principles. The first principle follows from the fact that most fundamental questions in neuroscience are cross-level questions: how do molecules and single neurons (at the level of microns) lead to whole-brain behavior and form mind (at the level of centimeters)? The major breakthroughs in the generation of neuroscientists will come from those that can think across levels. Thus, our first principle is that being trained at all levels, and having an intuition for the kind of research that occurs at every level, is essential. Second, given the complexity of the brain, next-generation neuroscientists will need equal familiarity with experimental and computational approaches. And third, familiarity with cutting-edge experimental methods, both from within and outside neuroscience, is crucial. The course teaching style was designed based on the principle that ﬁrst-hand experience with experiments (both wet and computational), is the best way to learn.
The following lecture modules and labs run concurrently:
Module 1: Cellular neurophysiology. Introduction to the fundamentals of how neurons operate physiologically. Excitability and function in neural tissue is covered in terms of energetics, single-cell dynamics, and synaptic transmission and plasticity. The module uses three current themes in neuroscience: monitoring function (cell-level measurements), dynamics of single neurons and circuits (firing properties and signal dynamics), and activity-dependent plasticity (synaptic plasticity and other forms of change on a time scale of seconds and longer). The biophysical principles that are shared across many or all animals are emphasized.
Module 2: Anatomy, connectomics, and organizational principles of brain structure. Principles of brain organization at the cellular and circuit level. Modern technologies in the new science of connectomics, in particular methods for analyzing and reconstructing microcircuits, are described. Emphasis is also placed on evolutionary principles and on selected example structures, including the hippocampus and cerebellum.
Module 3: Principles of circuits and neural dynamics. This module covers the neural mechanisms, mathematical basis, and computational properties of different categories of neural dynamics: persistent activity, oscillations, synchrony, and sequences. The mathematical dynamical principles underlying each of these types of dynamics is discussed, followed by discussion of papers from the primary literature that show example data that could be classified as one of these types of dynamics. Discussion centers on the circuit mechanisms that could underly these types of dynamics, the different types of computations such dynamics could support, and an examination of the evidence that these types of dynamics are in fact found in nervous systems.
Module 4: Neural development and plasticity. Introduction to developmental neuroscience, primarily using vertebrate model systems. Covers cellular events that are progressive (neurogenesis, neuronal migration, synaptogenesis, dendritic elaboration, axon elongation) and those that are regressive (cell death, synapse elimination, dendritic pruning, axon retraction) and the molecular mechanisms that underlie these phenomena. Discussion of temporal and spatial mechanisms of cell fate patterning, neuronal migration and axon guidance, establishment of topographic maps; the role of experience in shaping the developing brain, and the neural underpinnings of developmental critical periods; and structural and functional plasticity in the adult brain, including traditionally developmental events (e.g. neurogenesis, neuronal migration, dendritic remodeling) which continue throughout life.
Module 5: Molecular neuroscience and genetic model systems. An introduction to fundamental principles of molecular biology and their application to neuroscience. Lectures provide background on core principles of gene structure and expression, and discuss specific molecular techniques for monitoring and manipulating gene expression and defining circuit connectivity, as well as the use of genetically encoded tools to monitor, activate, or silence neural activity. This includes consideration of the advantages and weaknesses of specific genetic model systems and molecular methods. In the final class, students work in groups to present and critically evaluate papers from the primary literature that provide examples of molecular genetic approaches to fundamental questions in neuroscience.
Module 6: Low level sensory systems: Neural representation and coding. Principles of neural representation, neural encoding, and neural decoding, together with modern data analysis methods for understanding these processes. We cover the concepts of receptive field, spike-triggered averaging, linear and linear-non-linear models of receptive fields, plus modern nonlinear covariance-analysis-based methods. Experimenter decoding methods, including entropy and mutual information, signal and noise correlation, analysis of synergy and redundancy, and ROC analysis. Strengths and limitations of these methods are explored by applying them to real experimental data. Different representation schemes, including firing rates, synchrony, phase-of-firing, and spike timing representations are presented, and then discussed with respect to data from the literature on active sensing.
Module 7: Principles of systems and cognitive science. This module covers modern conceptual and theoretical frameworks used to understand and interpret data in systems and cognitive neuroscience. Particular attention is paid to decision-making and learning and memory. The module covers decision-making dynamics, including generalized drift-diffusion models, as well as Bayesian decision theory. The module then turns to principles of reinforcement learning, including Markov Decision Processes, value iteration, policy iteration, Temporal Difference learning, actor-critic architectures, and eligibility traces.
Module 8: High level sensory systems: Perception and attention. This module begins with a thorough review of visual pathways, from the retina, to subcortical nuclei, to primary visual cortex, to extrastriate cortex and the organization of higher order visual areas. It then continues with a review and substantial discussion of the selective attention literature, including space-based vs. object-based selection mechanisms, their neural basis, the effects of attention in sensory processing, attentional control systems in frontal and parietal cortex, and a discussion of attentional deficits, such as visuo-spatial hemineglect, in humans and animals models.
Module 9: Motor control and sequential action. Basic principles of organization of the motor system, ranging from muscle contraction, to spinal cord anatomy and control of reflexes, to motor cortex and premotor cortex and the many changing views of the cortical control of movement. The module also covers the basics of the parietal sensory-motor areas and their relationship to motor and premotor cortex, as well the oculomotor system, an especially revealing model system for sensory-motor integration. Modern work on computational models of motor control is then reviewed, including Bayesian approaches to motor control and state-based models of motor effectors and their control by primary motor cortices.
Module 10: Memory. An overview of learning and memory systems. Different and complementary learning systems and types of memory; comparison of hippocampal-dependent learning versus cortical-dependent learning. Proposed computational roles for hippocampus including pattern separation and pattern completion. Effects of lesions of different parts, or the whole, of the medial temporal lobe in both humans and animals. Mechanisms underlying forgetting (e.g., interference, contextual drift, retrieval-induced forgetting). The effects of sleep on learning and memory. Semantic-episodic interactions -- constructive memory and schemas. Prefrontal contributions to learning and memory.
Module 11: Learning and motivation. This module covers both computational and experimental approaches to learning and motivation. The module begins by reviewing experimental data from animals, beginning with classical conditioning, and uses these as a platform to think about the computational problem of prediction learning, how it might be solved, and possible neural substrates of prediction learning in the brain (dopamine). A host of popular algorithms from the computer science literature on reinforcement learning are covered and implemented by the students. Behavioral and fMRI studies of learning in humans are then considered, including modern cutting-edge computational techniques to predict trial-by-trial behavior; these methods are used to compare model-based versus model-free learning. We then consider motivation and addiction, exploration versus exploitation, and habitual versus goal-directed learning.
Module 12: Executive function and decision making. The module begins by considering decision-making, from both a computational and experimental point of view, covering both perceptual decision-making as well as reward-based decision-making, including expected value versus expected utility, prospect theory, and neural correlates of economic value. The module then turns to working memory, from both a psychological and neurophysiological point of view, and compares the relative roles of prefrontal cortex and posterior cortices. This leads to the concepts of executive function and cognitive control and the role of prefrontal cortex in these processes. Computational models of control and putative roles of neuromodulation (norepinephrine, dopamine) are considered. The module concludes by covering planning, problem solving, sequential decision-making, and performance monitoring, including conflict monitoring and anterior cingulate cortex.
Lab 1: Intracellular recording and synaptic potentials. The crayfish slow flexor muscle is used as a prep for recording intracellularly, using sharp electrodes, while stimulating afferent fibers. Students go from dissection to studying synaptic potentials.
Lab 2: Short-term synaptic plasticity. Having practiced the crayfish slow flexor muscle prep in lab 1, students now use it to study the difference between facilitation, potentiation, and augmentation in synaptic transmission.
Lab 3: In vivo structural plasticity. Students learn to implant cranial windows in mice and to use a two-photon microscope to image through them. Students image barrel cortex of transgenic mice expressing YFP in a mosaic pattern under a Thy-1 promoter; they do so once before, and once after, environmental enrichment, and look for changes in dendritic structure.
Lab 4: Circuit tracing. Students learn to use a trans-synaptic tracer, attenuated pseudorabies virus (PRV) expressing GFP, to trace pathways afferent to mouse cerebellum. Students obtain both epifluorescence and confocal images of brain slices.
Lab 5: Field potential measurements of LTP/LTD in hippocampal slices. Students prepare hippocampal slices and take field potential recordings while stimulating Schaffer collaterals in order to study LTP and LTD.
Lab 6: Intracellular patch recording in hippocampus. Students learn to perform whole-cell patch recordings in hippocampal slices, and perform a variety of studies using this preparation.
Lab 7: Optogenetics. Students use transgenic flies that express channelrhodopsin (ChR2) in sugar taste-sensitive neurons to compare action potential responses of these neurons to delivery of sucrose and delivery of blue light. In parallel, the proboscis extension reflex (PER) is measured, and students compare PER in response to sucrose and in response to blue light.
Lab 8: Computational decision-making dynamics. Students learn to simulate winner-take-all dynamics for decision-making and learn to numerically simulate the dynamical evolution of probability distributions of solutions (Fokker-Planck equations). Students are charged with exploring which biologically-plausible parameter changes in their Fokker-Planck model of 2-alternative-forced-choice decision-making could best explain recent experimental data on muscimol inactivation-induced changes.
Lab 9: “Connectionist” neural networks. Students learn to train multi-layer perceptron (or “backprop”) models, and use them to train these networks in a variety of student-chosen problems.
Lab 10: Measures of functional connectivity: spectral analyses of coherence and Granger causality. Using a dataset from multielectrode, multi-area recordings from monkeys performing a selective attention task, students learn to calculate power and coherence using both autoregressive modeling and modern multi-taper methods, as well as Granger causality spectra. Students analyze these data to determine whether the brain’s attention network in fronto-parietal cortex provides feedback signals to visual cortex to enhance sensory processing.
Lab 11: Models of muscle synergies. Students learn to use, and then compare, modern blind source separation techniques, including principal component analysis, non-negative matrix factorization, factor analysis, and independent component analysis, to study to what extent one can recover muscle synergies given recorded muscle activity
Lab 12: Mathematical models of memory. Students use an interactive version of the Temporal Context Model of memory recall that allows them to adjust parameters and explore how these parameters affect the model’s predictions, as well as to compare how well these predictions compare to an experimental data set.
Lab 13: Covertly measuring memory retrieval using multi-voxel pattern analysis. Students observe a real-time demonstration of fMRI data collection, and then use pre-collected data from a full memory retrieval experiment to learn how to use modern multivariate pattern classification algorithms.
Lab 14: Cognitive control. Students apply modern dense-scalp EEG recordings in an experiment to assess evidence regarding conflict-related negativity. Students compare their results to existing data and models in the literature.
APC/MAT 351 Topics in Mathematical Modeling - Mathematical Neuroscience This course combines modeling with applied math methods including PDE, probability, stochastic ODE, dynamical systems, cells as electrical circuits, Hodgkin-Huxely equation describing spikes in single neurons & bursting neurons (e.g., breathing, heartbeat, other rhythms), propagation of action potentials, reaction-diffusion equations, Hopfield-Grossberg neural nets, leaky accumulator models, drift-diffusion models, information theoretic approaches to analysis of neural spike trains.
NEU/MOL 408/PSY 404 Cellular and Systems Neuroscience A survey of fundamental principles in neurobiology at the biophysical, cellular, and system levels. Lectures will address the basis of the action potential, synaptic transmission, sensory physiology and motor control, development of the central nervous system, synaptic plasticity, and disease states. A central theme will be the understanding of systems phenomena in terms of cellular mechanisms (can be used as a first course in neuroscience for entering graduate students in Neuroscience who are coming from a different field and are not yet ready for the core curriculum).
MOL 431 Advanced Topics in Developmental Neurobiology Contemporary approaches to the study of neural development, emphasizing genetic and molecular techniques. Topics include generation, patterning, differentiation, migration and survival of neurons and glia, axon growth and guidance, target selection, synapse formation/elimination, activity-dependent remodeling of connectivity, and the relationship between neural development and behavior. Reading will be mainly from the primary literature with textbook reading provided for background. Classroom participation is required.
NEU/MOL 437/537 Computational Neuroscience Introduction to the biophysics of nerve cells and synapses, and the mathematical descriptions of neurons and neural networks. How do networks of neurons represent information, and how do they compute with it? The course will survey computational modeling and data analysis methods for neuroscience. Representation of visual information, navigation through space, short-term memory and decision-making will be some of the issues considered from a mathematical/computational viewpoint.
MOL 508 Advanced Topics in Neurobiology This course will focus on original scientific literature and class discussion with readings that center on major problems and current research in neuroscience.
MOL 510 Introduction to Biological Dynamics Designed for students in the biological sciences, this course focuses on the application of mathematical methods to biological problems. Intended to provide a basic grounding in mathematical modeling and data analysis for students who might not have pursued further study in mathematics. Topics include differential equations, linear algebra, difference equations, and probability. Each topic will have a lecture component and computer laboratory component. Students will work extensively with the computing package Matlab. No previous computing experience necessary.
NEU/PSY 511 Neuroscience seminar series: Current Issues in Neuroscience and Behavior Advanced seminar that reflects current research on brain and behavior.
NEU/PSY 330 Introduction to Connectionist Models: Bridging Between Brain and Mind A fundamental goal of cognitive neuroscience is to understand how psychological functions such as attention, memory, language, and decision-making arise from computations performed by assemblies of neurons in the brain. This course will provide an introduction to the use of connectionist models (also known as neural network or parallel distributed processing models) as a tool for exploring how psychological functions are implemented in the brain, and how they go awry in patients with brain damage.
PSY/NEU 336 The Diversity of Brains The premise of this seminar is that an understanding of the neural basis of behavior can be gained by examining species-typical behaviors. Each animal species has evolved neural solutions to specific problems posed to them by their environment. The course will focus primarily on forebrain mechanisms in mammals, highlighting the unique environmental problems that a species must solve and the ways in which the brains of these animals implement their solutions. Some example model systems include prey capture by bats, monogamy and aggression in voles, and eye gaze processing by primates.
PSY/NEU 338 Animal learning and decision making – psychological, computational and neural perspectives Seminar designed to expose students to a modern, integrative view of animal learning phenomena from experimental psychology, through the lens of computational models and current neuroscientific knowledge. At the psychological level we will concentrate on classical and instrumental conditioning. Computationally, we will view these as exemplars of prediction learning and action selection, the pillars of reinforcement learning. Neurally, we will focus on the roles of dopamine and the basal ganglia at the systems level. Students will see how the study of animal decision making can inform us about the computations that take place in the brain.
PSY 407 Developmental Neuroscience An analysis of cellular processes and regulatory factors that underlie vertebrate brain development and the development of behavior. Topics include: neurogenesis, neuronal migration, cell death, synapse formation, dendritic differentiation, as well as the influences of neurotransmitters, hormones, trophic factors and experience on developmental processes and behavior. In addition, conditions that induce abnormal brain development, and potentially result in the development of psychopathology, will be considered.
PSY/NEU 410 Depression: From Neuron to Clinic This course focuses on clinical depression, utilizing it as a model topic for scientific discourse. This topic is ideal for this purpose because it intersects a broad range of issues. The course focuses on a neurobiological approach to this personally and societally important subject. Topics range from the molecular to the clinical.
PSY/NEU/MOL 415 Advanced Topics in Learning & Memory: Cellular and Molecular Mechanisms Seminar designed to expose students to current research on the cellular and molecular basis of learning and memory, providing an up-to-date analysis of what is and is not known about the neurobiology of learning and memory. We begin with a review of the model systems used to study learning and memory, including an analysis of the translational validity of certain model systems. We then deal with different forms of plasticity (synaptic and structural) as they pertain to learning and memory during development and adulthood. Finally, we apply some of these findings to evaluate the current status of research on aging and Alzheimer's.
PSY/NEU 416 Brain Imaging in Cognitive Neuroscience Research This course will provide an introduction for advanced students on the use of functional brain imaging in cognitive neuroscience research. The first third of the course will cover the foundations of brain imaging in neurophysiology, imaging physics, experimental design, and image analysis. The rest of the course will be an examination of innovations in experimental design and methods of analysis that have opened new areas of cognitive neuroscience to inquiry using functional brain imaging.
PSY/NEU 516: The Neural Basis of Goal-Directed Behavior A fundamental property of human action is its orientation toward specific desired outcomes or goals. Understanding the computations & neural mechanisms underlying this goal-directedness is a central challenge for both psychology and neuroscience. We'll review major theories characterizing the role of goals in behavior, from cognitive, social & developmental psychology, animal behavior research and artificial intelligence. Having established this conceptual context, we'll review a wide range of neuroscientific data to sketch out the neural substrates of goal-directed behavior, considering the neural basis of goal evaluation, selection, representation & pursuit.
PSY 591A Ethical Issues in Scientific Research Examination of issues in the responsible conduct of scientific research, including the definition of scientific misconduct, mentoring, authorship, peer review, grant practices, use of humans and of animals as subjects, ownership of data, and conflict of interest. Class will consist primarily of the discussion of cases. Required for all first and second year graduate students in the Department of Psychology. Open to other graduate students.
Other Courses of Interest to Neuroscience Graduate Students
APC 503 Analytical Techniques/Differential Equations
APC 514 Biological Dynamics
CHE 514 Molecular and Biomolecular Imaging
CHM 545/MOL512 Magnetic Resonance in Chemical Biology and Neuroscience
CHV/NEU 510 Graduate Seminar in Neuroethics
COS 402 Artificial Intelligence
COS 429 Computer Vision
COS 487 Theory of Computation
EEB 502/3 Fundamental Concepts in Ecology, Evolution, and Behavior
NEU 593 Magnetic Resonance Imaging
MAE 541/APC541 Applied Dynamical Systems
MAE 546 Optimal Control and Estimation
MOL 504 Cellular Biochemistry
MOL 506 Molecular Biology of Eukaryotes
MOL 507 Developmental Biology
MOL 510 Introduction to Biological Dynamics
MOL 515 Methods and Logic in Quantitative Biology
MOL 561 Scientific Integrity
PHY 561/2 Biophysics
PSY 543 Research Seminar in Cognitive Psychology