The Connectomics team at the Allen Institute's brain science accelerator uses large scale anatomical methods to measure connectivity within the brain at a very large scale with single–cell resolution. There are two major projects within the department that operate at different scales. Electron microscopy (EM) connectomics attempts to measure connectivity and morphology of cells with single synapse resolution across millimeters of brain. Axonal connectomics uses light microscopy methods to reconstruct large caliber single axons as they course the brain’s white matter, at the scale of centimeter, in primate and human brains. Both projects also characterize and classify cell types based on their morphology and work on integrating these morphological classifications with other modalities such as transcriptomics and single-cell physiology, both in vivo and in vitro.

While the EM connectomics has formed the bulk of our research over the past decade, the newer Axonal Connectomics project aims to apply connectomics approaches to larger brains, with the complete human brain as the ultimate goal. While the EM approach can reasonably be applied to the entire mouse brain in the coming year, the required data sizes make it impossible to do EM imaging of entire human brain. With lower resolution, however, it is an achievable goal to map the majority of large (myelinated) axons in the coming years. Given that never has a single axon been traced from source to target through the human white matter, a new approach is clearly needed.

With a combination of antibody staining, tissue clearing, expansion, and light-sheet microscopy, we have begun mapping the courses of projection axons in large brains at high resolution. The resultant three-dimensional data sets are readily segmented both by humans and with machine-learning approaches, similar to those used with electron microscopy. We are currently working to scale-up our centimeter-scale data sets to increasingly large portions of larger brains, including the human.

‍