Leveraging Artificial Intelligence and Multi-Omics Data to Predict Opioid Addiction

Study Summary

The primary goal of this proposal is to validate a novel genomic and microbiome predictive model that may be used to assess a person's risk of developing opioid use disorder (OUD). The following will be tested: (1) MODUS (Measuring risk for Opioid use Disorder Using SNPs), which is a genomic panel consisting of a set number of proven single nucleotide polymorphisms (SNP) that utilizes machine learning to determine an individual's risk; and (2) MICROUD (MICRObiome for Opioid Use Disorder), which will be a novel microbiome prediction panel for OUD risk. MODUS and MICROUD will be developed using existing public datasets with genomic and microbiome data (e.g., All of Us, Human Microbiome Project). During development of these predictive models, in parallel, an external prospective validation cohort will be recruited consisting of subjects from the University of California, San Diego, Veteran Affairs of San Diego, and Veteran Affairs of Palo Alto (each site with separate IRB). The hypothesis is that MODUS and MICROUD will have high predictive potential for identifying high risk patients for OUD.

Primary Outcome

risk of opioid use disorder based on genomics, social determinants of health, microbiome, and other clinical data. Patients with OUD will be enrolled and compared to healthy controls using data from external datasets such as NIH All of Us Research Program. Diagnosis of opioid use disorder will be based on the DSM-5 criteria for opioid use disorder

Conditions Studied

Study Locations (1)