| 日 時 | 2022年07月26日(火) 16:00 より 17:00 まで |
|---|---|
| 講演者 | Marc M. Takeno, Ph.D. |
| 講演者所属 | Senior Research Associate, Allen Institute for Brain Science, Seattle, WA, USA |
| 場 所 | Online (Zoom) |
| お問い合わせ先 | 窪田芳之（世話人） yoshiy@nips.ac.jp |
| 要旨 |
On behalf of a multi-institute team, I will present highlights from a functional connectomics dataset that combines physiology and ultrastructural anatomy in the same volume of mouse visual cortex. This dataset is the largest to date that combines functional calcium imaging and ultrastructural connectivity in the same cells. The responses of an estimated 75,000 neurons from primary visual cortex and three higher visual areas were recorded from, using calcium imaging, while the mouse viewed natural movies and parametric stimuli. Electron microscopy data of the same volume was automatically segmented, and more than 200,000 neuronal and non-neuronal cells were reconstructed, and 524 million synapses automatically identified, from the same tissue. Manual proofreading of some cells from this volume yielded complete dendritic arbors and local and inter-areal axonal projections. Of the 601 neurons that were proofread in the publicly available dataset, 78 excitatory neurons have axons and dendrites proofread to their maximal extension within the EM volume. Axon arbor length ranges from 2.5 to almost 19 mm, with a mean number of 714 synapses per axon (range: 192 – 1893). For 104 inhibitory neurons that were proofread for false merges, the number of output synapses ranged from 40 to 10,051. As an example, one completely proofread Layer 2/3 pyramidal cell has 3441 inputs and 1398 outputs identified, with 1239 synaptic partners. The dense reconstruction and scale of the dataset allow for analysis between groups of neurons (motifs), layer-layer connectivity, and functional and anatomical analysis of feedforward and feedback connections that are not possible with other modalities and methods. This dataset is released as an open access resource at www.microns-explorer.org with analysis tools for programmatic data access, and a web user interface to view morphologies and data. I will also highlight aspects of method development that were crucial to producing this dataset, work in progress, and future directions for our research group using large-scale electron microscopy. Relevant links: https://www.microns-explorer.org/ https://www.biorxiv.org/content/10.1101/2021.07.28.454025v2 |
