Unveiling and Steering Connectome Organization with Interpretable Latent Variables

TL;DR

This paper introduces a framework that combines connectomics and representation learning to uncover low-dimensional, interpretable organizational principles of neural connectomes, enabling controlled manipulation of connectome subgraphs.

Contribution

It presents a novel method integrating subgraph extraction, generative modeling, and explainability to interpret and manipulate brain connectome structures.

Findings

Effective graph reconstruction demonstrated

Latent codes can be manipulated to generate connectomes with specific properties

Provides insights into structural features linked to function

Abstract

The brain's intricate connectome, a blueprint for its function, presents immense complexity, yet it arises from a compact genetic code, hinting at underlying low-dimensional organizational principles. This work bridges connectomics and representation learning to uncover these principles. We propose a framework that combines subgraph extraction from the Drosophila connectome, FlyWire, with a generative model to derive interpretable low-dimensional representations of neural circuitry. Crucially, an explainability module links these latent dimensions to specific structural features, offering insights into their functional relevance. We validate our approach by demonstrating effective graph reconstruction and, significantly, the ability to manipulate these latent codes to controllably generate connectome subgraphs with predefined properties. This research offers a novel tool for understanding brain architecture and a potential avenue for designing bio-inspired artificial neural networks.