Recent technological advances now enable in-depth assessment of blood based ‘omic’ markers at scale and can provide important insights into the aetiology of metabolic and other diseases. Existing studies are limited by their design, size, analytical strategy and single disease focus. We combine large-scale genomic with metabolomic and proteomic data to define the genetic architecture of thousands of ‘omic’ traits. We integrate data from different platforms and studies to maximise power for the identification of genetic variants associated with life-long differences in metabolite and protein levels and enable investigation of their relevance for disease aetiology and prediction, with a specific focus on type 2 diabetes and related traits.