Computational Modeling
Building on our collaboration with Wilfrid Rall on the first computer models of brain neurons (see Rall Archive), our lab has pioneered applying realistic computational modeling methods to experimental data, to reveal mechanisms of information processing in dendritic spines, dendritic trees, and cortical microcircuits.
A Hodgkin-Huxley like action potential in cortical dendrites
We used this data to extend the modeling approach to brain neurons as exemplified in mitral and granule cells, resulting in the prediction and confirmation of a dendrodendritic microcircuit for lateral inhibition (see Computational Modeling above, and the Wilfrid Rall Archive below). The figure illustrates the analysis of the functional properties of cortical dendrites under the assumptions of generation of a Hodgkin-Huxley like action potential antidromically invading the axon, soma, and either active or passive dendrites (Rall and Shepherd, 1968). This was the first exploration of a Hodgkin-Huxley like action potential in a compartmental neuron model.
Biophysical model of Ca spread in a dendritic spine
Another computational study has analyzed the spread of calcium within and out of the head of a dendrodendritic granule cell spine (see figure).
The narrow neck limits the spread, supporting the concept that the spine head can act as a semi-independent integrative input-output unit (Woolf and Greer, 1991).Logic operations from interactions between active dendritic spines
A global model of the initial stage of odor processing in the olfactory bulb
The video begins by growing a system of glomeruli (round balls) attached to apical mitral cell dendrites. Mitral cell lateral dendrites (purple) grow out in the external plexiform layer to encircle the olfactory bulb and interact with the dendrites of granule cells to gate the activation of the mitral cells coming from the glomeruli and apical dendrites activated by natural odor stimuli. (Migliore et al, 2014).