Summary Statistics for Genome-Wide Structural Variant Associations and Rare SV–Protein Level Associations Underlying "The Biomedical Landscape of Genomic Structural Variation in the Qatari Population" paper

Summary Statistics for Genome-Wide Structural Variant Associations and Rare SV–Protein Level Associations Underlying “The Biomedical Landscape of Genomic Structural Variation in the Qatari Population” paper These files contain full summary statistics for (1) genome-wide structural variant (SV) association analyses with quantitative traits and (2) rare SV association analyses with plasma protein levels in the Qatari population. The data accompany the manuscript “The Biomedical Landscape of Genomic Structural Variation in the Qatari Population”. Contents sv_gwas_traitX.zip Compressed archive of per-trait association results for genome-wide SV–phenotype analyses. Each file inside the archive is an EPACTS .txt file for one of 42 clinically relevant quantitative traits collected at Qatar Biobank (QBB). Results are provided for SVs with allele frequency (AF) > 0.1% and size ≤ 2 Mb, across 6,141 QBB participants. sv_pwas_proteinY.gz Tab-compressed file of .assoc outputs for rare SV–protein level association analyses (PWAS). Contains summary statistics for 61,898 rare SVs (AF < 1%) tested against 1,305 plasma proteins in 2,935 QBB participants. All files are summary-level association statistics; no individual-level genotype or phenotype data are included. Methods overview For the genome-wide SV–trait analyses, we tested SVs (AF > 0.1%, size ≤ 2 Mb) for association with 42 quantitative phenotypes measured at QBB. Traits were transformed using a rank-based inverse normal transformation prior to analysis. Association testing was performed with the q.emmax test in EPACTS v3.4.2 using default parameters and an AF cutoff of 0.1%. A kinship matrix for the 6,141 individuals was generated from SNP data using the EPACTS make-kin function and supplied to q.emmax. Age and sex were included as covariates. For the rare SV–protein level analyses, we assessed rare SVs (AF < 1%) for association with inverse-normal-transformed levels of 1,305 proteins. Analyses were conducted using RVTests v2.1.0. To account for relatedness among the 2,935 QBB participants, we generated a kinship matrix with the vcf2kinship function. We then applied a linear mixed model (FastLMM) in RVTests to obtain summary statistics for each SV–protein pair, adjusting for sex, age, and the first three genetic principal components. Multiple testing was controlled using a genome-wide significance threshold of P < 6.18 × 10⁻¹⁰ (0.05 / (61,898 SVs × 1,305 proteins)).