NeuroGraphBench: Interacting with Drosophila Connectomes at Scale for Exploring the Functional Logic of Neural Circuits
Year: 2026
Type: preprint
Abstract: Drosophila connectomic datasets provide increasingly comprehensive maps of neuronal morphology and synaptic connectivity, offering an unprecedented opportunity to explore the structural organization of its neural circuits. This calls for designing automated tools to interact with connectomic datasets at scale for efficiently exploring structural features embedded in the vast amounts of data. Yet the central challenge remains the understanding of the functional logic of neural circuits. In order to formulate a connectome-grounded hypothesis for what a circuit computes, it is critical to (i) characterize the objects in the natural environment in which brain circuits operate, and (ii) formulate how brain circuits represent and process the defined objects. To develop and demonstrate a methodology for these requirements, we focus on the Drosophila looming-evoked escape pathway. We modeled the trajectory of looming objects that are on a collision course (direct-hits) or pass by the fly (near-misses): their projected images on the retina can be characterized by the subtended solid angle and center elevation. By analyzing the sub-neuronal structure and retinotopic organization of the looming pathway embedded in connectomic datasets, we devised a pathway abstraction, and mapped its structural elements into a connectome-grounded functional model. We demonstrate that (i) the expanding boundary of a looming object activates an annulus of LPLC2 neurons, and (ii) the Giant Fiber reads out this population code through its two dendritic branches, modeled as separate units whose outputs are combined to drive its axon. It is this functional separation of the dendritic branches that gives the model access to the center elevation, separating direct-hits from near-misses well before the solid angle of the two trajectories diverge. To accelerate the connectomic analysis step, we developed a Python toolset with an agentic, code-free workspace interface called NeuroGraphBench (NGB). NGB provides four composable morphology-analysis primitives and an AI Agent that composes them to interactively respond to natural-language queries aided by visualization on an interactive 3D canvas. Thus, NGB automates tedious and repetitive tasks to enable faster and scalable connectomic exploration, keeping human reasoning, instead of writing code, at the center of an open-ended research inquiry.
Authors: Aurel A. Lazar, Shashwat Shukla, Yiyin Zhou
Institutions: Columbia University, Fordham University
Language: English
Subfield: Cellular and Molecular Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
Funder: National Science Foundation
OpenAlex ID: W7204449981