ABSTRACT This application is being submitted in response to Notice of Special Interest (NOSI) NOT-CA-24-032. Certain populations have been traditionally underrepresented in cancer research and therefore the benefits of cancer discoveries are more limited for these populations. Understanding informative variability (heterogeneity) across individuals is the key to prediction, and therefore to address underrepresentation it is critical to better understand heterogeneity within underrepresented groups. Cancers occur due to dysfunctional biological processes. We hypothesize that many important sources of within-group heterogeneity will influence gene function and biology and will be observed in patterns of gene expression (the transcriptome). Transcriptomes sum the biological effects of lifestyle, genetics and environmental exposures on gene expression and provide a molecular platform well-suited to explore tumor heterogeneities that may originate from many different factors. Further, tumor expression has been shown to have utility for predicting clinical risk and outcomes. Hence, transcriptomes provide an attractive way to understand both the origins and consequences of tumor heterogeneity. We will perform characterization of tumor transcriptomes to understand heterogeneity within three underrepresented populations: cancer patients living in rural areas, those of lower socio- economic status, and those of Hispanic ethnicity. Deep multi-dimensional characterization will be determined using the novel SPECTRA method developed by the Camp lab. Multiple independent quantitative transcriptome variables will be derived that describe gene expression variability within each group. A common limitation of molecular tumor studies is a paucity of companion epidemiologic data, necessary to identify potential avenues for intervention and modification of risk. In addition to data collected in the parent study, we will use record-linkage to the unique and powerful Utah Population Database (UPDB) to add individual- and area-level epidemiologic variables. For each of the three focus groups, we will construct rich datasets that will include demographic and lifestyle characteristics, clinical and prognostic variables, sociodemographic metrics, measures of comorbidity, healthcare access and environmental exposures. We will identify associations between state-of-the-art transcriptome variables and these elements to determine possible causes and consequences of tumor heterogeneities observed within each group. Findings from this project will narrow the knowledge gap by increasing our understanding of within-group tumor heterogeneities and has the potential to provide new avenues and opportunities to advance equity in cancer prevention, control and outcomes in these underrepresented populations.