ENIGMA- Addiction: Pooling of Existing Datasets to Identify Brain and Genetic Correlates of Addiction, Next Steps

PROJECT SUMMARY The ENIGMA consortium was established to investigate brain structure, function, and disease by combining genomic and neuroimaging datasets from multiple sites. Its cost-effective goal is to maximize statistical power and the yield from existing datasets through very large data pooling efforts. This goal has considerable importance in light of growing concerns over the rigor and reproducibility of many neuroimaging and genomic findings. ENIGMA has had many successes (including a genome wide association study on cortical thickness with over 51,556 participants, Grasby et al., 2021 Science), and a number of working groups have been established which leverage the standardized multi-site ENIGMA preprocessing pipelines and analytic methods to focus on the neurobiology of specific diseases. This application’s PI co-created the ENIGMA Addiction working group which now has access to datasets representing over 24,000 participants contributed by 103 PI members from 71 different institutions in 16 countries spanning 6 continents. There have been 27 peer reviewed publications since the start of the initial R01 with another 8 papers currently under review. Building on our initial funding (R01 DA047119), we propose to further expand the Addiction working group to support the largest international resource of its kind for the scientific community. Dataset users are encouraged to propose and lead their own projects. Curation of the shared database by the ENIGMA Addiction team according to standardized protocols for harmonization and quality control is a cost-effective means of maximizing heavy investments in data collection and overcomes an insurmountable obstacle for younger investigators and investigators from low or middle income countries. ENIGMA Addiction has approved 43 projects, 16 of which are completed and 27 are ongoing, including a full 30 projects led by investigators from outside of the central research team at the University of Vermont. These projects investigate topics ranging from sex differences to early markers of addiction risk to machine learning classification. Research projects can be performed via guest access to a high performance supercomputer hosted at the University of Vermont or through COINSTAC, a federated cloud- based meta-analytic pipeline. Analyses have been completed on thousands of pooled participants with structural MRI, DTI, resting state fMRI or genetic data and we have now expanded to include task-based fMRI. During the next phase of funding, the team at UVM will engage in two main analytic efforts whose success is favored by the large database afforded by this working group, namely comprehensive interrogation of psychiatric comorbidity by joining efforts with other large ENIGMA working groups and, second, identification of biologically meaningful sub-groups within the population of problematic substance users with distinct etiology.