While much is known about in vitro pharmacology of psychoactive drugs at individual receptors , the integrated effects on neurons--let alone whole animals--is far from clear. Recently, we have developed an exciting multi-electrode array assay to measure neuronal firing patterns in acute brain slices. In a single experiment we can monitor ~100 individual neurons over hours while applying precise pharmacological perturbations. This platform enables us to test for differential neurophysiological effects across related compounds, and combining these recordings with “opto-tagging” of neurons with channelrhodopsin2 permits identification of subpopulations of neurons via genetic phenotype and circuitry.
We, Elyssa Margolis at UCSF and Matt O'Meara at the University of Michigan are looking to co-supervise a postdoc with a computational or statistical background and interest in electrophysiology to design experiments and analyze data from this large-scale electrophysiology assay. The first set of experiments will be designed to improve our understanding of the neuronal mechanisms underlying responses to opioid compounds and how they generate pain relief, tolerance, dependence, and addiction .
Computationally, the approach will be build on recent developments in deep learning to construct differentiable simulators for the receptor perturbation, signaling mechanism(s), and observation process. We can then fit these models by explicitly or implicitly backpropagating the discrepancy with the collected data back to the model parameters. Modeling full data generation process makes the analysis more rigorous so it can scale with the complexity of the experiments. Further, by assigning priors to the parameters of interest, this approach enables designing maximally informative experiments .
 (Lyu, et al., Nature, 2019)
Ultra-large library docking for discovering new chemotypes
 (Margolis and Fields 2017; Margolis et al., 2014; Fields and Margolis 2015)
 (Foster, et al., AI-STATS 2020) A Unified Stochastic Gradient Approach to Designing Bayesian-Optimal Experiments
Behavior is energized by motivations and shaped by either positive (rewarding) or negative (punishing) outcomes. Opioids, both endogenous and exogenous, act in the CNS to powerfully modulate motivation and the experiences of either reward or punishment. Understanding how opioids contribute to these processes requires identifying the responsible neural circuits and then determining their synaptic actions on the component neurons of these circuits. Various project opportunities are available to advance our understanding of opioid receptor neuropharmacology in the neural circuits that underly opioid reinforcement and analgesia, and the changes in synaptic function and opioid receptor function following drug self administration, drug dependence, or injury. We also have ongoing projects investigating to the role of endogenous opioids in modulating alcohol consumption. Most projects integrate a variety of behavioral/in vivo, ex vivo, and/or anatomical techniques.
Expertise in at least one of the following techniques is preferred: rodent behavior/analysis, electrophysiology, immunocytochemistry, and/or stereotaxic surgery. Proficiency in programming is a plus. Successful candidate(s) will be expected to lead a project as well as be able to work as part of a highly collaborative lab team. The laboratory is located in the Sandler Neurosciences Building on the Mission Bay Campus, providing access to the UCSF Neuroscience community including the UCSF Weill Institute for Neurosciences, the Center for Integrated Neuroscience, and the Kavli Institute for Fundamental Neuroscience.
The University of California San Francisco is an Equal Opportunity/Affirmative Action Employer. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, age, protected veteran or disabled status, or genetic information.