Skip to content
DraconiticResearch
Search Sources API
203,234 studies31 sourcesUpdated 49 min ago
Peer-reviewedCC BY-NC-NDNeuron

Inferring brain-wide interactions using data-constrained recurrent neural network models

  • Matthew G. Perich1,2
  • Charlotte Arlt3,4
  • Sofia Soares3,4
  • Siyan Zhou3,2
  • Manuel Beirán5,6,2
  • Aaron S. Andalman7,8,6
  • Tyler Benster9,2
  • Megan E. Young10,8,6
  • Clayton P. Mosher11,2
  • Juri Minxha12,4
  • Eugene Carter10,2
  • Ueli Rutishauser12
4 more authors
  • Peter H. Rudebeck10
  • Christopher D. Harvey3
  • Karl A. Deisseroth13
  • Kanaka Rajan14
14 affiliations
  1. 1Département des neurosciences, Université de Montréal, Montréal, QC, Canada; Quebec Artificial Intelligence Institute (Mila), Montréal, QC, Canada
  2. 2Icahn School of Medicine at Mount Sinai, New York, NY, USA
  3. 3Department of Neurobiology, Harvard Medical School, Boston, MA 02115, USA
  4. 4Harvard Medical School, Boston, MA, USA
  5. 5Zuckerman Mind Brain Behavior Institute, Columbia University, New York, NY, USA
  6. 6Cedars-Sinai Medical Center, Los Angeles, CA, USA
  7. 7Department of Bioengineering, Stanford University, Stanford, CA, USA
  8. 8California Institute of Technology, Pasadena, CA
  9. 9Neurosciences Graduate Program, Stanford University, Stanford, CA, USA
  10. 10Nash Family Department of Neuroscience and Friedman Brain Institute, Icahn School of Medicine at Mount Sinai, New York, NY, USA
  11. 11Nash Family Department of Neuroscience and Friedman Brain Institute, Icahn School of Medicine at Mount Sinai, New York, NY, USA; Cedars-Sinai Medical Center, Los Angeles, CA, USA
  12. 12Cedars-Sinai Medical Center, Los Angeles, CA, USA; California Institute of Technology, Pasadena, CA, USA
  13. 13Department of Bioengineering, Stanford University, Stanford, CA, USA; Department of Psychiatry and Behavioral Sciences, Stanford University, Stanford, CA, USA; Howard Hughes Medical Institute, Stanford University, Stanford, CA, USA
  14. 14Department of Neurobiology, Harvard Medical School, Boston, MA 02115, USA; Kempner Institute for Natural and Artificial Intelligence, Harvard University, Boston, MA 02134, USA. Electronic address: [email protected]
Open at sourceDOICite

Abstract

Behavior arises from the coordinated activity across anatomically and functionally distinct brain regions. Modern experimental tools allow unprecedented access to large neural populations spanning many interacting regions brain-wide. Yet, understanding such large-scale datasets necessitates robust, scalable computational models to extract meaningful features of inter-region communication and principled theories to interpret those features. Here, we introduce current-based decomposition (CURBD), an approach for inferring brain-wide interactions using data-constrained recurrent neural network models that autonomously produce dynamics consistent with experimentally obtained neural data. CURBD leverages the functional interactions inferred from such models to reveal directional currents between multiple brain regions simultaneously. We first show that CURBD accurately isolates inter-region currents in simulated, ground-truth networks with known connectivity and dynamics. We then apply CURBD to multi-region neural recordings obtained from many species-larval zebrafish, mice, macaques, and humans-to demonstrate the widespread applicability of CURBD in untangling brain-wide interactions and inter-area communication principles underlying behavior.

Related

  1. Graph neural network approach for inferring neuronal connectivity
    A. Gupta, C. L. Neveu, J. H. Byrne
    bioRxivPreprintCC BYFull textNeurosciencebioRxiv
  2. Inferring the latent network of pairwise mutualistic preferences from observed plant-pollinator interactions
    L. Federici, E. Matechou, I. Iacopini
    bioRxivPreprintCC BY-NC-NDFull textEcologybioRxiv
  3. A physics-informed neural network framework for inferring infinitesimal phase response curves from finite-pulse circadian recordings
    D. Won, G. Hwang, J. K. Kim
    bioRxivPreprintCC BY-NC-NDFull textBioinformatics & computational biologybioRxiv
  4. Attractor dynamics underlie sequence working memory in data-constrained recurrent neural networks
    M. Pals, J. H. Macke
    bioRxivPreprintCC BYFull textNeurosciencebioRxiv
  5. NeuralRNN: a unified framework for recurrent neural network methods in cognitive neuroscience
    H. Chen, N. Ding
    bioRxivPreprintCC BYFull textNeurosciencebioRxiv

Cite

BibTeXRIS
BibTeX
@article{perich2026inferring,
  title = {{Inferring brain-wide interactions using data-constrained recurrent neural network models}},
  author = {Matthew G. Perich and Charlotte Arlt and Sofia Soares and Siyan Zhou and Manuel Beirán and Aaron S. Andalman and Tyler Benster and Megan E. Young and Clayton P. Mosher and Juri Minxha and Eugene Carter and Ueli Rutishauser and Peter H. Rudebeck and Christopher D. Harvey and Karl A. Deisseroth and Kanaka Rajan},
  journal = {Neuron},
  year = {2026},
  month = aug,
  doi = {10.1016/j.neuron.2026.07.016},
  url = {https://doi.org/10.1016/j.neuron.2026.07.016},
  note = {CC BY-NC-ND},
}
RIS
TY  - JOUR
TI  - Inferring brain-wide interactions using data-constrained recurrent neural network models
AU  - Matthew G. Perich
AU  - Charlotte Arlt
AU  - Sofia Soares
AU  - Siyan Zhou
AU  - Manuel Beirán
AU  - Aaron S. Andalman
AU  - Tyler Benster
AU  - Megan E. Young
AU  - Clayton P. Mosher
AU  - Juri Minxha
AU  - Eugene Carter
AU  - Ueli Rutishauser
AU  - Peter H. Rudebeck
AU  - Christopher D. Harvey
AU  - Karl A. Deisseroth
AU  - Kanaka Rajan
PY  - 2026
DA  - 2026/08/01
JO  - Neuron
DO  - 10.1016/j.neuron.2026.07.016
UR  - https://doi.org/10.1016/j.neuron.2026.07.016
AB  - Behavior arises from the coordinated activity across anatomically and functionally distinct brain regions. Modern experimental tools allow unprecedented access to large neural populations spanning many interacting regions brain-wide. Yet, understanding such large-scale datasets necessitates robust, scalable computational models to extract meaningful features of inter-region communication and principled theories to interpret those features. Here, we introduce current-based decomposition (CURBD), an approach for inferring brain-wide interactions using data-constrained recurrent neural network models that autonomously produce dynamics consistent with experimentally obtained neural data. CURBD leverages the functional interactions inferred from such models to reveal directional currents between multiple brain regions simultaneously. We first show that CURBD accurately isolates inter-region currents in simulated, ground-truth networks with known connectivity and dynamics. We then apply CURBD to multi-region neural recordings obtained from many species-larval zebrafish, mice, macaques, and humans-to demonstrate the widespread applicability of CURBD in untangling brain-wide interactions and inter-area communication principles underlying behavior.
SN  - 0896-6273
SN  - 1097-4199
ER  -