Two theoretical postdoctoral positions and 1-2 collaborative experiment-theory positions are available in the laboratory of Dr. Mark Goldman at the University of California at Davis.  The lab works on a broad range of problems in computational neuroscience ranging from neural coding to dynamics and plasticity of single neurons and networks.  Immediate funding is available for a range of projects related to working memory, neural integration, motor learning, and synaptic plasticity as described below.  The postdoctoral candidate also would have flexibility to work on a range of issues of his or her choosing.  Candidates are expected to have strong training in an analytically rigorous discipline such as theoretical neuroscience, physics, mathematics, computer science, or engineering.  The postdoctoral candidate will have ample opportunity to interact within the vibrant computational and systems neuroscience communities at UC Davis and the greater San Francisco Bay Area.    

     Candidates should send a CV, brief statement of previous research and future research interests, and contact information for three references to: Mark Goldman, msgoldman@ucdavis.edu.

Specific positions with funding include:

1) Theoretical positions:  Dynamics of memory and motor-related neural activity

     Challenging the attractor picture of working memory.  In the traditional attractor picture of working memory, memory storage results from positive feedback processes that lead to the formation of self-sustained attractors.  In one project, we are exploring how functionally feedforward, rather than feedback, network architectures can generate flexible codes for storing memories and producing a broad range of input-output transformations.  In a second project, we are utilizing methods from engineering control theory to show how balanced cortical networks can utilize negative feedback to stabilize persistent patterns of neural activity. 

     Multi-scale modeling of neural integration.  The oculomotor neural integrator, which transforms eye-velocity encoding motor commands into eye-position encoding commands, is a model system for understanding the mathematical integration of inputs and the maintenance of memory-storing activity.  We seek to determine the respective roles of cellular and circuit mechanisms of memory storage in this system.  Multi-scale models, from ion channels to behavior, will be generated based upon electrophysiological and optical imaging recordings and optogenetic manipulations performed in our experimental collaborators’ laboratories. 

     Context-dependent memory storage.  Recent experiments suggest that the oculomotor neural integrator functions in a context-dependent manner, producing spatially distinct activity patterns depending upon whether eye movements are being made in the context of rapid changes in gaze (saccades) versus smooth tracking of moving objects.  We seek to determine the circuit mechanisms underlying this context-dependent activity and to propose general frameworks, applicable to both oculomotor and cortical memory systems, for how multiple context-dependent inputs can be separately stored in a single network.

     Role of the granule cell layer in cerebellar motor learning.  The eye movement system provides a highly tractable setting for studying motor learning because it is well-characterized experimentally and has fewer degrees of freedom than more complicated movement systems.  In collaboration with whole-circuit optical imaging and optogenetic perturbation experiments in the Aksay laboratory at Cornell Medical University and genetic manipulations, electrophysiological recording, and optogenetic perturbations in Jennifer Raymond’s laboratory at Stanford University, we are modeling the neural dynamics and coding of cerebellar granule neurons and their relation to Purkinje cell firing and the plasticity of eye movement behaviors. 

2) Experimental positions:

Development, dynamics, and plasticity of neural networks:  In collaboration with Kim McAllister (kmcallister@ucdavis.edu), we are seeking to understand the learning rules underlying development and learning in neural networks.  We will use cutting-edge technology in patterned substrates, optogenetics, and uncaging in a novel long-term imaging assay for synapse dynamics that allows recording and single synapse or single cell manipulation of neuronal and network activity.  These cultured networks will be used to directly test central tenets of Hebbian, spike-timing dependent, and homeostatic learning over development.  The applicant will perform these imaging experiments as well as apply theoretical models to better understand and predict our results.

Dynamics of memory and motor-related neural activity: There are also possibilities for joint experimental-theoretical work with Emre Aksay’s laboratory at Weill Medical College of Cornell University (in New York City) on either oculomotor memory storage or cerebellar motor learning, using in vivo two-photon optical recording and stimulation to dissect circuit function and plasticity.  Interested candidates should contact Dr. Aksay at ema2004@med.cornell.edu.