The present study reinforces recent discoveries that show the inherent power of adding phenotypes to the genetics prediction model in order to increase accuracy8. Several group-specific PRS were identified as distinguishing between the continued AUD and the remitted AUD group. For example, while the EA male model includes PRS related to aggression, the EA female model includes PRS related to depression. Interestingly, both EA group models for the prediction of AUD remission include neuroticism PRS and years of education PRS. Our findings are in line with recently published studies showing PRS association with disorder outcomes in depression56, schizophrenia57, and alcohol-related phenotypes58. These findings highlight the potential of PRS collections to predict the course of development and recovery from diseases. PRS collections representing genetic fingerprint of various phenotypes allow embodying complex diseases with multiple domains. Notably, our results demonstrate the significance of using genetic data over self-report to identify self-ancestry, which increases the genetic homogeneity of the groups, leading to higher prediction scores.
