Finding Patterns Within the Connectome

Methods

Summary

The data in the connectome represents one of the most complex datasets that has been undertaken for study. The connectome is a multi-layered dynamic system. For example, the brain can be thought of as a collection of discrete subunits, each of which conducts a specific task - such as understanding speech, or facial expressions, or interpreting tastes. The wiring between these areas, and how their information flow is coordinated, is one layer of the connectome. The neurons in each subunit are organized quite specifically (and also dynamically, responding to external stimuli and reconfiguring, a phenomenon we call "plasticity"). For example, the cortex, a layer of the mammalian brain which conducts the most complex processing, is organized in a layers of millions of copies of column-like structures. Within each cortical column, neurons are organized similarly, while their interconnections are quite specific and unique. Knowing these interconnections comprises another layer of our connnectome. Further detail and another level is the biochemical level, where neurons communicate with other neurons via synapses, which are spaces between neurons that are connected chemically or electrically. A variety of neurotransmitters carry out communication at this level. Another level of detail is the cell types themselves. Human brains, indeed all mammalian brains, are comprised of a variety of neural cell types - over 200 distinct cell types (more information at http://neuromorpho.org/). A connectomic dataset keeps track of all this information -- the conformation and configuration of neural cell types, all synaptic interconnections, the specific neurotransmitters used for communication and the varying concentrations of these neurotransmitters, neural organization in microstructures such as the cortical column, the connections between the microstructures, the neural network structure within the subunits, and the wiring between subunits.

And you thought it was going to be complicated, didn't you?

To begin this Herculean task we must first choose a small part of it. We will be looking at small biological neural networks (BNN) to discover their signaling dynamics, in both simulation (using the software Neuron) and also experimentally, using multi-electrode array (MEA) analysis. MEAs connect neurons to an electronic interface, allowing direct signal data from a neuron to be captured. This interesting set of experiments will allow us to study the actual signals between neurons that are in the process of storing information.

Do you want to know more? Ask us!