Sex-based Network-Specific Differences in Connectomes: A Krakencoder-Based Analysis

2026-06-15 · Source: Computer Vision and Pattern Recognition · Field: Science & Research — Life Sciences & Biology, Artificial Intelligence & Machine Learning, Mathematics & Computational Sciences · Depth: Expert, quick

Summary

A study utilizing the Krakencoder simulation framework analyzed how deficiencies in one brain connectome modality propagate to another. Researchers examined structural and functional connectomes from 702 healthy participants in the Human Connectome Project, assessing the impact of each of the Yeo-7 functional networks. Seven scenarios involved removing a single network, and cross-modal prediction perturbations were quantified using KL divergence on eigenvalue spectra, Frobenius norm, and Wasserstein distance. The Default Mode Network consistently produced the largest perturbations, while the Somatomotor network yielded the smallest across all metrics and prediction directions. Sex differences in network-level perturbation signatures were subtle, achieving a maximum accuracy of 66.09% in sex classification under network-removal conditions. In contrast, connectomes predicted from intact inputs showed substantially higher sex classification accuracy, reaching up to 84.76%, confirming they retain considerably more sex-discriminative information.

Key takeaway

For research scientists investigating brain network resilience or sex-based neurological differences, you should recognize the Default Mode Network's disproportionate impact on connectome integrity. Your models predicting sex-specific information from perturbed connectomes will likely show significantly reduced accuracy (e.g., 66.09% vs. 84.76% for intact inputs), indicating that full connectomes are crucial for retaining such discriminative data.

Key insights

Krakencoder analysis shows Default Mode Network perturbations most impact connectome integrity and sex-discriminative information, which is largely retained in intact predictions.

Principles

• Default Mode Network is critical for connectome integrity.
• Intact connectomes retain more sex-discriminative data.
• Network-specific perturbations have varying impacts.

Method

The Krakencoder framework simulates connectome modality propagation by removing single Yeo-7 networks. Perturbations are quantified via KL divergence, Frobenius norm, and Wasserstein distance, followed by evaluating sex-specific information persistence in predicted connectomes.

Topics

• Krakencoder
• Brain Connectomes
• Default Mode Network
• Network Perturbation
• Sex Differences
• Human Connectome Project

Best for: AI Scientist, Research Scientist

Related on AIssential

• They Copied a Fly's Brain Into a Computer. It Started Walking. Nobody Taught It How — Biological brain wiring can inherently produce complex behaviors without explicit training or optimization.
• Network ontology transcript annotation identifies genetic signals underlying sex determination — Silver ion treatment in cannabis perturbs phytohormone signaling, inducing sex reversal by altering cell wall and embryogenic tissue genes.
• EssentialGIN: a new approach for gene essentiality prediction based on graph isomorphism neural networks — Graph isomorphism networks, by integrating biological data and preserving network topology, significantly improve essential gene prediction accuracy.
• Hybrid knowledge- and data-driven modelling for robust spike detection and sorting in human C-fiber microneurography — A hybrid pipeline improves C-fiber spike sorting in microneurography by combining physiological knowledge with machine learning for robust detection.
• Harnessing cortical geometry, wiring, and function as inductive biases for recurrent neural networks — Cortical geometry, wiring, and function serve as powerful inductive biases for more effective recurrent neural networks.
• Community Concealment from Unsupervised Graph Learning-Based Clustering — Community concealment in GNNs depends on boundary connectivity and feature similarity, mitigated by targeted perturbations.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Computer Vision and Pattern Recognition.