Gordon Shepherd
Keynote lecture
Will talk about: The olfactory system as a model for building the tools of neuroinformatics
Gordon M. Shepherd introduced the olfactory bulb as a model system for analysing the synaptic organization of the brain. He was a pioneer with Wilfrid Rall, a founder of computational neuroscience, in building the first compartmental models of brain neurons. This work led to a microcircuit model of the olfactory bulb, one of the first of a brain region, a model which has been tested and refined by his lab and many others over the years. He also introduced activity-dependent methods for revealing how odor molecules creat odor maps in the olfactory bulb, which has led to the current analysis of the maps with a wide range of methodologies by many laboratories.
Shepherd obtained an M.D. at Harvard Medical School and a D.Phil. at Oxford. After postdoctoral work at NIH and the Karolinska Institute, he came to Yale in 1967, where he is Professor of Neurobiology. He has had visiting positions at the University of Pennsylvania, College de France, Ecole Normale Superieure, Pasteur Institute, and Oxford University. He has served as editor in chief of the Journal of Neurophysiology and the Journal of Neuroscience. His books include The Synaptic Organization of the Brain (5th ed.), Neurobiology (3rd ed.), Foundations of the Neuron Doctrine, Creating Modern Neuroscience: The Revolutionary 1950s, Handbook of Brain Microcircuits (ed with S. Grillner), and Neurogastronomy.
In addition to his experimental and computational work, Shepherd was among the founders of the field of neuroinformatics. He and his colleagues Perry Miller and Michael Hines built one of first neuroinformatics websites, called SenseLab, containing a suite of 8 databases serving research into the synaptic organization of neurons and neural microcircuits, and into the family of olfactory receptor genes, the largest family in the genome.
SenseLab is an interoperable set of databases to support the development of a comprehensive framework for the neural mechanisms underlying microcircuit function in the brain. The olfactory system is providing an attractive model for this purpose. ORDB archives over 14,000 chemoreceptor genes which transduce chemosensory stimuli into receptor cell responses. OdorDB archives over 200 odor molecules and the receptors with which they interact, while a new database, OdorModelDB, is being developed to support molecular modeling of the odor-receptor interactions. OdorMapDB contains spatial patterns elicited by the different odors in the glomerularlayer of the olfactory bulb. Research on the mechanisms of neuronal processing in the olfactory bulb is supported by a subset of four databases. CellPropDB contains membrane properties expressed by different neurons in the olfactory bulb and over 30 brain regions. This is expanded in NeuronDB to show the detailed expression patterns within different dendritic, somatic, and axonal compartments, which enables the integration within a compartment to be analyzed. After identifying an integrative motif in a part of a neuron, a unique multi-domain search tool enables testing for the generality of that motif across all brain regions. Quantitative data for the membrane properties in the different neuronal compartments are contained inModelDB, which now comprises over 700 published and curated models. These are searchable by multiple criteria such as brain region, authors, functional operations, and modeling program. The models range from individual ion conductances and neurotransmitter receptors through dendrites and differenttypes of neurons to multineuronal ensembles. The ensembles are collected in MicrocircuitDB, which is being developed to identify principles of microcircuit organization that apply broadly across all brain regions. Finally, BrainPharmDB is being developed to extend the analysis of normal function to nervous disorders such as Alzheimer's. These databases and tools are enhancing research in these areas, as well as identifying broader principles of neuronal organization. This will be illustrated by a model of neuronal processing underlying the perception of smell and flavor, and a new hypothesis of forebrain evolutionfrom three-layer to six-layer cortex.
