UK Biobank Genetic Analyses Enhancement

During the next 18 months, it is intended to measure about 600,000 genetic markers in the DNA extracted from blood samples that have already been collected from each of the 500,000 participants in UK Biobank. When these “genotype” measurements are combined with whole genome sequence data from a few thousand or tens of thousands of UK individuals, it will be possible to “impute” (i.e. estimate) very many more genetic variants in the region of the DNA adjacent to the variants that have been measured. The combination of these detailed genotyping data with the extensive range of known biochemical risk factors that are currently being measured in blood and urine samples from the UK Biobank participants, along with the detailed information from questionnaires and physical measurements conducted at the initial assessment visits and from linkage to health records about the development of disease during long-term follow-up, will make UK Biobank uniquely rich as a resource for researchers from all areas of health to conduct studies of the relevance of genes to disease rapidly and cost-effectively. Hence, these detailed genotype data will facilitate research that harnesses the full power of UK Biobank to help understand the causes of many different diseases.

During the next 2 years, all 500,000 participants in UK Biobank will be genotyped. Dense “biobank” genotyping chips have recently been developed that combine measurement of about 250,000 SNPs in a whole-genome array, about 200,000 markers covering copy number variation, loss of function, insertions and deletions, and previously identified risk factor or disease associations, and about 150,000 exome markers which cover a high proportion of non-synonymous coding variants with allele frequency over 0.02%. These “biobank” genotyping chips provide a powerful tool for enabling genome-wide interrogation of common variants, and a scaffold that allows extension to lower frequency and rarer variants by imputation from reference sequence data sets (e.g. 1000Genomes, UK10K, GoT2D). Availability of these genotype data on the whole UK Biobank cohort would allow inference of haplotype phase at the genotyped SNPs at higher accuracy than is typical with whole-genome association studies (GWAS), using novel methods that exploit haplotype sharing within populations. The exome-focused component would allow exploration of phenotypic effects of a large number of functional coding variants, which would benefit from the data on a wide range of traits that are available on the participants in UK Biobank. Improvements in inference of phase, and the subsequent imputation, would increase the statistical power for associating genetic variants with biomarkers, phenotypes and disease outcomes. These genotyping data would remain of value even when UK Biobank participants start to be sequenced as they can be used to validate the sequencing data and to impute even more detailed genetic information for the whole cohort. Combination of such rich genotyping data with the extensive range of known biochemical risk factors currently being measured (about 50 analytes that are related to cardiovascular disease, cancer, metabolic and renal disease, and other conditions), along with the detailed phenotyping from the baseline assessment visit (which includes information about prevalent disease) and data on the development of disease during follow-up from linkage to health records, will make UK Biobank uniquely rich as a resource for scientists from all areas of health research to conduct genetic association studies rapidly and cost-effectively (since they would only need to request data rather than fund genotyping). Hence, acquisition of dense genotype data in all of the UK Biobank participants will facilitate analyses that harness the full power of the resource.