Jeff W. Lichtman: The Past, Present and Future of Connectomics

Despite enormous effort and interest, the brain remains the most poorly understood aspect of human biology. How behavior, thoughts, and, more generally, consciousness are instantiated in the brain's physicality is a genuinely hard problem.  The challenge stems from the fact that the brain is a network. The billions of cells (“neurons”) that comprise a brain are each connected by microscopic synapses to many thousands of other neurons. The individual wires that convey the signals from one neuron to another are often far-flung, sometimes spanning meters.  Moreover, neurons vary widely in structure, function, and connectivity. Indeed, the brain has far more cell types than the rest of the body combined. The goal of connectomics is to describe this wiring diagram in its entirety so that every neuron and every synaptic connection between neurons (there are trillions of synapses in a human brain) are mapped. I will describe how this goal emerged, where we are in obtaining such maps, and, thanks to AI, what the near future of connectomics looks like. 

Jeff Lichtman is Jeremy R. Knowles Professor of Molecular and Cellular Biology at Harvard. He received an AB from Bowdoin (1973), and an M.D. and Ph.D. from Washington University (1980) where he worked for 30 years before moving to Cambridge in 2004. He is a member of the Center for Brain Science. Lichtman’s research interest revolves around the question of how mammalian brain circuits are physically altered by experiences, especially in early life.  He has focused on the dramatic re-wiring of neural connections that takes place in early postnatal development when animals are doing most of their learning. This work has required development of techniques such as “Brainbow” transgenic mice to visualize neural connections and monitor how they are altered over time.  Recently his efforts have focused on developing new electron microscopy methods to map the entire wiring diagram of the developing and adult brain.  This "connectomics" approach has as one of its aims uncovering the ways information is stored in neural networks.