Neuronal Diversity: Types and Teams
Recently, much interest has focused on questions about the number of different types of neurons in the brain, or in specific brain regions. Much of this work has focused on questions of molecular type. We are interested in questions about the functional differences between cell types at the level of intrinsic, synaptic, circuit and tuning properties. Using statistical models and techniques such as stimulus reconstruction we seek to determine how specific functions, such as stimulus encoding, are implemented by the different types of neurons in the networks performing these tasks. One element of this work has been focused on developing a public database of neuronal properties from the literature that we call NeuroElectro.org. Beyond differences across cell types, we have explored the functional consequences of cell-to-cell diversity. Within the population of cells of a given type we have shown that the cells display a range of properties and that this diversity can contribute to the ability of populations of cells to represent complex stimuli. We also have shown that in olfactory bulb mitral cells a significant fraction of this diversity is eliminated by blocking a single type of voltage-gated potassium channel. We are interested in the mechanisms that underlie this diversity and also in how differences in intrinsic properties relates to differences in synaptic connections. Olfactory Navigation: Smelling your way
Animals have a remarkable ability to navigate using odor cues. Whether its scent dogs tracking a lost child or pocket rats finding landmines, the olfactory navigation many abilities of animals far outstrip any technological solutions developed by humans. As part of an multi-PI NSF-funded project we are trying to determine the algorithms and mechanisms used by animals to solve olfactory navigation problems. By coupling newly developed methods for analyzing the behavior of mice, flies and other species, with clever approaches to monitoring turbulent flow of odors, we will identify the cues used by animals performing these tasks and also understand how these cues are integrated by brain networks. This project has its own web site at odornavigation.org. |
Neuronal Models: Stats + Maths
Neuronal activity can be modeled in (at least) two ways - using statistical models and mechanistic models. Statistical models simply predict the spiking of a neuron given the input and the recent history of spiking. Mechanistic models incorporate what we know about channels and membranes to create models that not only capture the behavior, but also connect in clear ways to the biological features of neurons such as channel properties and densities. These two types of models offer important advantages: statistical models offer insight about what activity encodes or represents, whereas biophysical models allow predictions to be made about perturbations to biological features. We seek to connect statistical to biophysical models in order to answer questions like "What do Kv1.2 channels contribute to a neuron's ability to encode stimuli?" Neuronal Reliability: What is Noise?
Neurons are unreliable. That is, neurons or populations of neurons that receive identical inputs many times over, do not respond identically time after time. Are biological devices just intrinsically unreliable? or are there advantages to this variability? What are the mechanisms of this variability and why is it enhanced in the brains of people with autism and also in mice with genetic mutations that mirror those seen in some patients with autism? Olfactory Plasticity: Are you Experienced?
The relatively simple and highly stereotyped circuitry of early stages of the olfactory system make it an ideal system in which to analyze experience-dependent plasticity. We are analyzing how the connectivity and function of olfactory bulb circuits are influenced by changes to an animal's odor environment and determining whether these effects of experience are restricted to specific periods of development. |