Connectomics Analysis
Rreverse-engineering the most complex processor in the universe could be much easier if we had a complete schematic of the computing elements and how they interconnect. Connectomics attempts to build this schematic from real neural systems using incredibly advanced neurotechnology. I'm interested in extracting understanding from the resulting schematics. What are the repeat patterns of connectivity? What are the canonical microcircuits? What are the types of computing elements?
Our Automatic Discovery of Cell Types project is an initial step in this journey, working with Konrad Kording We've put up an entire site for the project with code, data, etc.
The follow-on to that project was joint work with Srini Turaga at HHMI/Janelia on using continuous, latent-space models for uncovering network structure, and is still ongoing. We presented a poster at COSYNE in 2016.
publications & posters
Automatic discovery of cell types and microcircuitry from neural connectomics Eric Jonas and Konrad Kording. eLife, April 30 2015 paper [our site ]
Kernel Latent Space Models for understanding neural Connectomes Eric Jonas and Srini Turaga, COSYNE 2016 [poster pdf here]
Press
"How the Emerging Revolution in Neural Wiring Diagrams is About to Change Biology" - The Physics arXiv Blog