Next Week: CV-DL Seminar – Dean Geckt | Network motif analysis In a connectomics-based reconstructed mouse V1
1 view
Skip to first unread message
Naaman Kopty
May 25, 2026, 7:10:47 AM (6 days ago) May 25
to 'Google Groups' via CV-DL Seminar – University of Haifa
Dear all,
You are in shavited to attend the next talk in the CV-DL Seminar Series.
Dean Geckt
HUJI
HUJI
Sunday, May 31, 2026
11:15-12:30
Room 508, Amir Building
11:15-12:30
Room 508, Amir Building
Title:
Network motif analysis In a connectomics-based reconstructed mouse V1
Network motif analysis In a connectomics-based reconstructed mouse V1
Abstract:
This study investigates the role of specific connection patterns, or "motifs," in shaping the
function of complex neural networks. We analyzed a high-resolution connectome—a detailed
map of 1,351 neurons and 140,173 synapses—from the mouse visual cortex - MICrONs V1.
Our systematic search for the most statistically significant three-node subgraphs (triplets)
revealed a dominant pattern: the "bi-mutual Inhibitory-Excitatory-Excitatory (IEE) triplet." This
motif is characterized by one inhibitory neuron (I) that forms reciprocal, two-way connections
with two separate excitatory neurons (E1 and E2), where the two excitatory neurons (E1 and
E2) are not directly connected to each other. This motif appears far more frequently than
predicted by a random network model while maintaining the same proportion of excitatory to
inhibitory connections per node and an identical distribution of degrees.
We discovered that this IEE Bi-Mutual motif is not randomly placed, but possesses distinct
spatial and anatomical features:
● Spatial Organization: The inhibitory neurons are typically positioned below their
excitatory partners. Furthermore, the unconnected excitatory neurons are spatially closer
to each other in the network's depth dimension compared to unconnected E-nodes
outside the motif.
● Connection Anatomy: Inhibitory synapses strongly avoid targeting distal compartments,
such as the apical tuft.
● Multiple Contacts: There is a strong tendency for multiple excitatory and inhibitory
contacts at both the individual neuron and the triplet level. Notably, multiple excitatory
contacts tend to be asymmetric, with an asymmetry factor up to five.
These results conclusively demonstrate that the bi-mutual IEE motif is a highly structured
architectural feature with non-random spatial and connection properties. This structured
embedding suggests it may serve a critical function, potentially stabilizing or regulating network activity, similar to a signal "whitening" or "normalization/balancing" mechanism observed in other nervous systems. Future work will employ simulations and functional data (calcium data from the same V) to test which network dynamics and computational outcomes are best supported by this precise structural organization.
This study investigates the role of specific connection patterns, or "motifs," in shaping the
function of complex neural networks. We analyzed a high-resolution connectome—a detailed
map of 1,351 neurons and 140,173 synapses—from the mouse visual cortex - MICrONs V1.
Our systematic search for the most statistically significant three-node subgraphs (triplets)
revealed a dominant pattern: the "bi-mutual Inhibitory-Excitatory-Excitatory (IEE) triplet." This
motif is characterized by one inhibitory neuron (I) that forms reciprocal, two-way connections
with two separate excitatory neurons (E1 and E2), where the two excitatory neurons (E1 and
E2) are not directly connected to each other. This motif appears far more frequently than
predicted by a random network model while maintaining the same proportion of excitatory to
inhibitory connections per node and an identical distribution of degrees.
We discovered that this IEE Bi-Mutual motif is not randomly placed, but possesses distinct
spatial and anatomical features:
● Spatial Organization: The inhibitory neurons are typically positioned below their
excitatory partners. Furthermore, the unconnected excitatory neurons are spatially closer
to each other in the network's depth dimension compared to unconnected E-nodes
outside the motif.
● Connection Anatomy: Inhibitory synapses strongly avoid targeting distal compartments,
such as the apical tuft.
● Multiple Contacts: There is a strong tendency for multiple excitatory and inhibitory
contacts at both the individual neuron and the triplet level. Notably, multiple excitatory
contacts tend to be asymmetric, with an asymmetry factor up to five.
These results conclusively demonstrate that the bi-mutual IEE motif is a highly structured
architectural feature with non-random spatial and connection properties. This structured
embedding suggests it may serve a critical function, potentially stabilizing or regulating network activity, similar to a signal "whitening" or "normalization/balancing" mechanism observed in other nervous systems. Future work will employ simulations and functional data (calcium data from the same V) to test which network dynamics and computational outcomes are best supported by this precise structural organization.
Bio:
Dean Geckt is a PhD student in Computational Neuroscience at the Hebrew University of Jerusalem. His research focuses on bridging structure, function, and dynamics in neural circuits, specifically working with the MICrONS connectome dataset to uncover the functional role of network motifs.
Prior to his doctoral studies, Dean earned an M.Sc. in Computer Science with highest distinction from the University of Haifa. His master's thesis on code-switching strategies in human-machine dialogs was successfully published in Bilingualism: Language and Cognition. He also holds a B.Sc. in Computer Engineering from the Technion.
Dean Geckt is a PhD student in Computational Neuroscience at the Hebrew University of Jerusalem. His research focuses on bridging structure, function, and dynamics in neural circuits, specifically working with the MICrONS connectome dataset to uncover the functional role of network motifs.
Prior to his doctoral studies, Dean earned an M.Sc. in Computer Science with highest distinction from the University of Haifa. His master's thesis on code-switching strategies in human-machine dialogs was successfully published in Bilingualism: Language and Cognition. He also holds a B.Sc. in Computer Engineering from the Technion.
We look forward to seeing you there.
Reply all
Reply to author
Forward
0 new messages