Network-Based Discovery of Opioid Use Vulnerability in Rats Using the Bayesian Stochastic Block Model

Area of Science:

• Neuroscience
• Behavioral Science
• Data Science

Background:

• Opioid use disorder (OUD) is a major public health epidemic, affecting over 200,000 individuals annually in the U.S.
• Individual vulnerability to OUD varies significantly, but pre-clinical identification of these sub-populations is challenging.
• Complex multivariate measurements in animal models hinder the identification of distinct opioid vulnerability levels.

Purpose of the Study:

• To develop and apply a novel non-linear network-based data analysis workflow.
• To identify opioid use sub-populations based on behavioral traits.
• To assess the contribution of behavioral variables to opioid vulnerability and resilience.

Main Methods:

• Utilized seven key behavioral traits from over 400 heterogeneous stock rats across two locations.
• Integrated data, removed batch effects, and constructed a rat-rat similarity network based on behavioral patterns.
• Applied community detection using a Bayesian degree-corrected stochastic block model to uncover sub-populations.

Main Results:

• Identified three statistically distinct clusters representing behavioral sub-populations: vulnerable, resilient, and intermediate for heroin use, refraining, and seeking.
• Demonstrated how behavioral variables interact to define these distinct sub-populations.
• The analysis workflow successfully differentiated rats based on their behavioral responses to heroin.

Conclusions:

• A novel network-based approach effectively identifies distinct behavioral sub-populations related to opioid vulnerability.
• This methodology provides a powerful tool for pre-clinical identification of individuals at different risk levels for OUD.
• The developed open-source R package, mlsbm, facilitates the application of this analysis workflow.