The present study, therefore, aims to create an ML model, predicting future AUD remission among individuals who had met criteria for active DSM-5 AUD at their first interview but no longer meet criteria for current DSM-5 AUD at their next interview. It should be noted that remission is a complex, multidimensional process; this study focuses on the reduction of alcohol use and of AUD symptoms to subclinical levels, which is one component of remission20. We used longitudinal multidimensional data from COGA (e.g., clinical, electrophysiological, GWAS, demographics), including individuals of European Ancestry (EA) and African Ancestry (AA). COGA collects data and follows individuals with AUD, providing a unique opportunity to compare an individual’s AUD status over the development of their addiction and during their remission. Most importantly the diverse COGA data-enabled stratified analyses, increased group homogeneity, creating an individualized model, and discriminative key features for every group. To further increase group ancestry genetic homogeneity we calculated the models when ancestry was based on self-report and again based on genetics calculated with ancestral principal components (PCA)21. Our central hypothesis was that model
