CHANGER: A novel genetic reference from 902 unrelated Chilean exomes to enhance South American medical genetics and research.

BackgroundGenetic studies have disproportionately focused on populations of European ancestry, limiting the generalizability of allele-frequency references and genetic associations to underrepresented groups, including South American populations. This gap is particularly relevant for rare diseases and cancer, where accurate variant interpretation depends in part on appropriate population context. In addition, population-specific haplotype structure influence genome-wide association analyses and the portability of polygenic scores across ancestries. By focusing on the Chilean population, our work aims to bridge these gaps, providing more accurate reference data for genetic research and enhancing diagnostic and therapeutic strategies for underrepresented communities. MethodsWe aggregated exomes from seven Chilean cohorts and processed samples using a standardized best practices workflow, including population-scale quality control and relatedness filtering. Aggregated variants were annotated with VEP (including LOFTEE, REVEL, AlphaMissense, EVE, and CADD) and clinically classified using InterVar. Population structure was assessed with PCA/admixture. We compared overlap and allele frequencies against external references, performed gene-level variant burden analysis, and constructed phased haplotypes for ancestry association using linear regression with Bonferroni correction. Finally, we developed a dashboard with R shiny framework to enable gene-wise exploration of annotated variants. ResultsWe present CHANGER (Chilean Aggregated National Genomics Resource), a comprehensive aggregation of 902 unrelated Chilean exomes designed to create a detailed reference of genetic variation in Chile. By incorporating data from multiple cohorts, we identified 774,110 unique genetic variants, with an average of 42,601 aggregated variants per individual. We identified 132,363 novel variants, of which 31,470 were common within our cohort (frequency > 1 %). We provide variant annotation and direct comparisons with other publicly available general population references. Beyond variant discovery, CHANGER supports gene-level burden scans (61 Missense-Damaging depleted genes; 2 Loss-of-Function enriched genes) and a haplotype resource for ancestry association (51,171 haplotypes), enabling downstream interpretation, improved imputation, and more powerful genome-wide association analyses in Chileans. ConclusionsCHANGER provides a valuable resource for genetic research and a reference for local variant interpretation. Importantly, CHANGER follows a high-quality methodological baseline and an open, FAIR-aligned infrastructure designed to grow, increasing its value for discovery, imputation, and equitable clinical interpretation over time.