Fast state-space methods for inferring dendritic synaptic connectivity
We present fast methods for filtering voltage measurements and performing optimal inference of the location and strength of synaptic connections in large dendritic trees. Given noisy, subsampled voltage observations we develop fast l[subscript 1]-penalized regression methods for Kalman state-space m...
| Main Authors: | , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Springer US
2016
|
| Online Access: | http://hdl.handle.net/1721.1/105880 https://orcid.org/0000-0002-9256-6727 |
| _version_ | 1826205011626950656 |
|---|---|
| author | Pakman, Ari Smith, Carl Paninski, Liam Huggins, Jonathan H. |
| author2 | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory |
| author_facet | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Pakman, Ari Smith, Carl Paninski, Liam Huggins, Jonathan H. |
| author_sort | Pakman, Ari |
| collection | MIT |
| description | We present fast methods for filtering voltage measurements and performing optimal inference of the location and strength of synaptic connections in large dendritic trees. Given noisy, subsampled voltage observations we develop fast l[subscript 1]-penalized regression methods for Kalman state-space models of the neuron voltage dynamics. The value of the l[subscript 1]-penalty parameter is chosen using cross-validation or, for low signal-to-noise ratio, a Mallows’ C[subscript p]-like criterion. Using low-rank approximations, we reduce the inference runtime from cubic to linear in the number of dendritic compartments. We also present an alternative, fully Bayesian approach to the inference problem using a spike-and-slab prior. We illustrate our results with simulations on toy and real neuronal geometries. We consider observation schemes that either scan the dendritic geometry uniformly or measure linear combinations of voltages across several locations with random coefficients. For the latter, we show how to choose the coefficients to offset the correlation between successive measurements imposed by the neuron dynamics. This results in a “compressed sensing” observation scheme, with an important reduction in the number of measurements required to infer the synaptic weights. |
| first_indexed | 2024-09-23T13:05:21Z |
| format | Article |
| id | mit-1721.1/105880 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T13:05:21Z |
| publishDate | 2016 |
| publisher | Springer US |
| record_format | dspace |
| spelling | mit-1721.1/1058802022-10-01T12:58:08Z Fast state-space methods for inferring dendritic synaptic connectivity Pakman, Ari Smith, Carl Paninski, Liam Huggins, Jonathan H. Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Huggins, Jonathan H. We present fast methods for filtering voltage measurements and performing optimal inference of the location and strength of synaptic connections in large dendritic trees. Given noisy, subsampled voltage observations we develop fast l[subscript 1]-penalized regression methods for Kalman state-space models of the neuron voltage dynamics. The value of the l[subscript 1]-penalty parameter is chosen using cross-validation or, for low signal-to-noise ratio, a Mallows’ C[subscript p]-like criterion. Using low-rank approximations, we reduce the inference runtime from cubic to linear in the number of dendritic compartments. We also present an alternative, fully Bayesian approach to the inference problem using a spike-and-slab prior. We illustrate our results with simulations on toy and real neuronal geometries. We consider observation schemes that either scan the dendritic geometry uniformly or measure linear combinations of voltages across several locations with random coefficients. For the latter, we show how to choose the coefficients to offset the correlation between successive measurements imposed by the neuron dynamics. This results in a “compressed sensing” observation scheme, with an important reduction in the number of measurements required to infer the synaptic weights. National Science Foundation (U.S.) (CAREER Grant) McKnight Foundation (Scholar Award) National Science Foundation (U.S.) (Grant IIS-0904353) Columbia College. Rabi Scholars Program 2016-12-19T20:42:24Z 2016-12-19T20:42:24Z 2013-09 2013-07 2016-08-18T15:43:05Z Article http://purl.org/eprint/type/JournalArticle 0929-5313 1573-6873 http://hdl.handle.net/1721.1/105880 Pakman, Ari et al. “Fast State-Space Methods for Inferring Dendritic Synaptic Connectivity.” Journal of Computational Neuroscience 36.3 (2014): 415–443. https://orcid.org/0000-0002-9256-6727 en http://dx.doi.org/10.1007/s10827-013-0478-0 Journal of Computational Neuroscience Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. Springer Science+Business Media New York application/pdf Springer US Springer US |
| spellingShingle | Pakman, Ari Smith, Carl Paninski, Liam Huggins, Jonathan H. Fast state-space methods for inferring dendritic synaptic connectivity |
| title | Fast state-space methods for inferring dendritic synaptic connectivity |
| title_full | Fast state-space methods for inferring dendritic synaptic connectivity |
| title_fullStr | Fast state-space methods for inferring dendritic synaptic connectivity |
| title_full_unstemmed | Fast state-space methods for inferring dendritic synaptic connectivity |
| title_short | Fast state-space methods for inferring dendritic synaptic connectivity |
| title_sort | fast state space methods for inferring dendritic synaptic connectivity |
| url | http://hdl.handle.net/1721.1/105880 https://orcid.org/0000-0002-9256-6727 |
| work_keys_str_mv | AT pakmanari faststatespacemethodsforinferringdendriticsynapticconnectivity AT smithcarl faststatespacemethodsforinferringdendriticsynapticconnectivity AT paninskiliam faststatespacemethodsforinferringdendriticsynapticconnectivity AT hugginsjonathanh faststatespacemethodsforinferringdendriticsynapticconnectivity |