Genotyping Array Design and Data Quality Control in the Million Veteran Program

Abstract

The Million Veteran Program (MVP), initiated by the Department of Veterans Affairs (VA), aims to collect biosamples with consent from at least one million veterans. Presently, blood samples have been collected from over 800,000 enrolled participants. The size and diversity of the MVP cohort, as well as the availability of extensive VA electronic health records, make it a promising resource for precision medicine. MVP is conducting array-based genotyping to provide a genome-wide scan of the entire cohort, in parallel with whole-genome sequencing, methylation, and other 'omics assays. Here, we present the design and performance of the MVP 1.0 custom Axiom array, which was designed and developed as a single assay to be used across the multi-ethnic MVP cohort. A unified genetic quality-control analysis was developed and conducted on an initial tranche of 485,856 individuals, leading to a high-quality dataset of 459,777 unique individuals. 668,418 genetic markers passed quality control and showed high-quality genotypes not only on common variants but also on rare variants. We confirmed that, with non-European individuals making up nearly 30%, MVP's substantial ancestral diversity surpasses that of other large biobanks. We also demonstrated the quality of the MVP dataset by replicating established genetic associations with height in European Americans and African Americans ancestries. This current dataset has been made available to approved MVP researchers for genome-wide association studies and other downstream analyses. Further data releases will be available for analysis as recruitment at the VA continues and the cohort expands both in size and diversity.

Keywords: GWAS; Million Veteran Program; SNP array design; VA; biobank; clinical variants; genetic ancestry; genetic relatedness; genotype data; quality control.

Figure 1

Key MVP 1.0 Genotyping Array Modules The modules are divided into those shared with the Axiom Biobank Genotyping Array and those unique to the MVP 1.0 array, along with descriptions and counts of unique markers in each module. Counts represent the number of markers in the module, and markers can be in more than one module.

Figure 2

Quality-Control Assessments on the MVP Dataset after Performance of the Advanced Marker Quality Control Procedures (A) MAF distribution after sample QC filtering. The inset diagram shows the distribution for markers with a MAF below 1%. (B) Minor-allele discordance rates per MAF bin, based on intentionally duplicated samples. (C) Marker missingness rates per MAF bin, after sample QC filtering. (D) Comparison of MAFs between the EA subset of MVP (MVP-EUR) and the UK Biobank European subset (UKB-EUR). (E) Comparison of MAFs between MVP-EUR and the non-Finnish European subset of gnomAD (gnomAD-NFE).

Figure 3

Analysis of Genetic Ancestry in the MVP Dataset (A) Density plots of the total length of runs of homozygosity (ROH) per individual in each genetic-ancestry subgroup. Only the top five most common subgroups are shown. (B) Principal-component analysis of the 1000 Genomes Project phase 3 dataset with MVP samples projected onto principal components 1 and 2. (C) The number of MVP samples in each genetic-ancestry subgroup as inferred by ADMIXTURE percentages and our thresholds. For a single ancestry subgroup, such as MVP_GBR, the threshold is at least 80% inferred for that ancestry (e.g., MVP_GBR is GBR > 80%). For a pair of identified subgroups, the two ancestries must be at least 90% combined (e.g., MBP_GBR_YRI is GBR + YRI > 90%). MVP_OTHER includes all samples that had less than 80% ancestry aligned to any reference population and less than 90% combining any two populations. (D) Visualization of ancestry subgroups via Uniform Manifold Approximation Projection (UMAP).

Figure 4

GWAS of Height with MVP Cohort (A) Replication of the direction of effect for markers previously associated with height as annotated in the NHGRI-EBI GWAS Catalog in the MVP cohort of non-related European Americans (n = 291,609). Color coding denotes the genetic ancestry of the original cohort in which the markers were associated with height. (B) Same as (A) except with the MVP cohort of non-related African Americans (N = 73,190).