Abstract The purpose of the NIDA Animal Genetics Program is to identify genetic, genomic, epigenetic variants, physiology and brain functions that contribute to addiction-like behaviors, related behavioral endophenotypes, and behavioral comorbidities to substance use disorder. During the past four years, our multidisciplinary and highly collaborative consortium has been identifying gene variants that are associated with increased vulnerability to compulsive oxycodone use, tolerance to the analgesic effects of oxycodone, and development of withdrawal-induced hyperalgesia by performing the first GWAS using an advanced model of chronic intravenous oxycodone self-administration in N/NIH heterogeneous stock (HS). We have also created the first preclinical oxycodone biobank which enables researchers who do not have the resources to perform chronic intravenous self-administration or next-generation genome sequencing to perform advanced genetic, molecular, and cellular studies to further our understanding of the biological changes underlying addiction-like behaviors. While these efforts have been very successful in achieving the planned milestones, it has become clear that our project would benefit from an even larger sample size. In particular, increasing sample sizes lead to exponential rather than linear increase in the number of loci identified, and would allow us to identify sex- specific gene variants. Moreover, in the past four years there has been tremendous technological advances in behavioral and genetic analysis that can be leveraged to provide unprecedented access to identify the single nucleotide and structural variants that contribute to complex behavioral endophenotypes of high relevance to oxycodone use-disorders. The first goal of this competing renewal is to double the sample size of the current GWAS to increase the number of gene variants identified including sex-specific variants and meet the demands of the Biobank. The second goal is to use high-throughput behavioral phenotyping using markerless pose estimation based on machine learning with deep neural network to identify behavioral endophenotypes that can help predict and identify individuals with a resistant, mild, moderate, or severe phenotype of oxycodone addiction-like behaviors. The third goal is to use methodological improvements of the genetic analysis, including the analysis of structural variants and tandem repeats, as well as enhanced integration with gene expression data. The fourth goal is to strengthen the oxycodone biobank infrastructure. This project is likely to continue having a sustained and powerful impact on the field because it will provide an exponential increase in the number of genetic loci identified, eQTLs and PheWAS analysis related to addiction-like behavior; establish the first high-throughput behavioral motifs analysis of addiction-like behaviors using parallel video-recording and automated machine learning analysis; identify novel behavioral endopheno...