Genome Positioning System (GPS) Reference Data v3

Version 3.0 of the Genome Positioning System (GPS) reference dataset. This version updates the GPS ancestry reference framework by replacing the admixed American (AMR) reference from the 1000 Genomes Project with a Native American (NAT) reference derived from HGDP Native American populations. The resulting GPS reference space contains five reference groups: AFR, NAT, EAS, EUR, and SAS. The NAT reference includes HGDP Maya, Pima, Karitiana, Surui, and Colombian samples. AFR, EAS, EUR, and SAS reference samples are derived from the 1000 Genomes Project. This release contains the complete frozen reference package required to run the GPS pipeline, including: Variant inclusion lists PCA projection loadings and allele frequencies LD-pruned SNP list Reference population centroids Reference population covariance matrices Pooled reference covariance matrix Reference sample PC coordinates The reference PCA, centroids, and covariance matrices were recomputed using the updated AFR/NAT/EAS/EUR/SAS reference panel. GPS therefore uses a single fixed reference PCA space for projecting external cohorts and calculating Mahalanobis distances and affinity metrics. Version 3.0 supersedes version 2.0 for new GPS analyses. Version 2.0 remains available for reproducibility of analyses performed using the previous AFR/AMR/EAS/EUR/SAS reference framework. This dataset accompanies the Genome Positioning System (GPS) pipeline available on GitHub.