Abstract Adeno-associated virus (AAV)-based gene therapies have shown significant promise for the treatment of rare diseases with unmet medical needs. In recent years, however, multiple reports of treatment-related serious adverse events have been reported in several clinical trials. A recent U.S. Food and Drug Administration (FDA) advisory committee meeting highlighted the impact of high vector doses on hepatic and renal toxicities following AAV gene transfer. Hence, there is an urgent need to develop strategies to achieve therapeutic efficacy at lower vector dosage. To achieve such, a better understanding of AAV biology in different species is critical. During the previous funding cycle, we worked towards optimizing preclinical-to-clinical translation of AAV vectors, which has proven difficult due to incongruent vector transgene expression profiles in different species. To this end, our lab first developed a body of work based on structure-guided AAV evolution. Then, we developed a novel, cross-species evolution approach, wherein AAV libraries are cycled sequentially across mice, pigs and macaques. This approach yielded novel AAV variants that are highly potent and cross-species compatible (ccAAVs). While the potential for clinical translation at lower vector doses using ccAAVs is exciting, understanding the mechanism(s) underlying improved transduction efficiency is critical. Over the next 5 years, we propose to dissect structure-function-mechanism correlates of AAV transduction in mouse and monkey models. Specifically, we will carry out multi-trait mapping (tropism, transduction, epigenetics, transcription, immunogenicity) of ccAAVs and related mutants in normal mice, CRISPR-based editor mice and rhesus macaques. Specifically, we will probe the bold, new concept that AAV capsids can dictate epigenetic fate and transcriptional activity of vector genomes across different species. Thus, the overarching goal of this application is to better understand the structure-function-mechanism nexus in AAV biology to enable improved vector design.