Ancestry imbalance and population structure in openSNP, a public direct-to-consumer genomic resource - PLINK files

This repository contains the processed genetic datasets described in the manuscript:"Ancestry imbalance and population structure in openSNP, a public direct-to-consumer genomic resource" (currently under submission). The data originates from public sources, primarily openSNP, alongside reference panels from the Human Genome Diversity Project (HGDP) and the 1000 Genomes Project (1KGP). All data has been remapped to the GRCh37 (hg19) assembly. Data Processing Overview Data integrity checks and remapping were performed using the apriha/snps Python library. Standardization to VCF format was conducted via changlubio/GenomePrep, followed by conversion to PLINK binary format (.bed, .bim, .fam). For each dataset, two versions are provided: Unfiltered: Raw merged/processed data. Filtered: Quality-controlled files with a 5% threshold for missing genotypes (--geno 0.05) and missing individuals (--mind 0.05). File Structure 23andme.zip / AncestryDNA.zip / FamilyTreeDNA.zip: PLINK files for individual DTC companies. Includes both unfiltered and filtered versions. DTC_merged.zip: The merged dataset of the three companies above (6,118 individuals and 73,382 SNPs in the filtered version). full_dataset.zip: The final integrated dataset containing the merged DTC data combined with HGDP and 1KGP reference panels (9,376 individuals and 66,843 variants in the filtered version). Usage Notes Relatedness: No removal of related individuals was performed, as it was not required for the specific scope of the original analysis. Reference Genome: GRCh37 (hg19). Sorting: The final full dataset has been sorted by family and within-family IDs using the PLINK --indiv-sort command.