Plasma proteomic associations with genetics and health in the UK Biobank.
Level 3 - non-randomized controlled study
Large cohort observational proteogenomic mapping study
PubMed 37794186 · doi:10.1038/s41586-023-06592-6
What was done
Researchers from the UK Biobank Pharma Proteomics Project characterized the plasma proteomes of 54,219 UK Biobank participants across 2,923 proteins. They conducted protein quantitative trait locus (pQTL) mapping to identify genetic associations with plasma protein levels, performed ancestry-specific mapping in non-European individuals, evaluated epistatic effects, and assessed genetic proxies for therapeutic targets and disease phenotypes.
What was found
pQTL mapping across the 2,923 proteins identified 14,287 primary genetic associations, of which 81% were previously undescribed. The study identified extensive trans pQTL architectures across cytokine and complement networks, long-range epistatic interactions involving ABO blood group and FUT2 secretor status on gastrointestinal tissue-enriched proteins, and evaluated genetically proxied target effects for targets such as PCSK9 and COVID-19 susceptibility loci.
Why it matters
This resource markedly expands the mapped genetic architecture of the human plasma proteome, providing an open-access platform to accelerate biomarker discovery, disease prediction modeling, and therapeutic target prioritization.
Limits
The abstract does not report specific effect sizes, variance explained, or discrimination metrics for the health prediction models. The UK Biobank cohort is predominantly of European ancestry, and the assay measured a targeted subset of 2,923 proteins rather than the full human proteome.
Cited by
- supports The UK Biobank is measuring proteomic profiles across 500,000 participants at a cost of approximately $50 per person, having already profiled over 50,000 individuals.