Skip to content

UKBB PPP

The UK Biobank Pharma Proteomics Project (PPP)1 is an effort to characterize the plasma proteomes of a sample of UK Biobank participants. The PPP was funded and organized by a consortium consisting of the research arms of major pharmaceutical companies. The aim was to produce datasets suitable for drug-discovery research.

Motivation

Often, GWAS hits are mechanistically mysterious: it is unclear what biological pathway links a given genetic polymorphism with an associated phenotype. At a high level the PPP is motivated by the observation that since proteins are the fundamental biological effector molecules, they can help explain GWAS hits. If we observe that a GWAS hit for a phenotype is also associated with changed plasma levels of a protein, we may reason that the protein lies along the biological pathway leading to the GWAS hit.

Public Data

While access to individual-level PPP data requires UKBB accreditation, there are a number of powerful summary PPP datasets that have been publicly released.

pQTLs

A plasma pQTL for a protein is a genetic variation that has likely-causal association with variations in plasma levels of that protein. One can find such pQTLs by running a GWAS with the plasma level of the target protein as the phenotype, then fine-mapping the results. The original Sun et al. paper1 provides pQTLs for 2525 different proteins (see Supplementary Tables 9 and 10). These pQTLs are divided into cis-pQTLs (within 1Mb of the gene encoding the protein they affect), and trans-pQTLs.

pQTLs can be used in Mendelian randomization (MR). In pQTL-based MR, the pQTLs for a protein are instruments, the level of the protein is the exposure, and a phenotype is the outcome. The goal is causal inference about the effect of the protein on the phenotype. In a pQTL MR study, one must decide whether to exclusively use cis-pQTLs, or to use both trans- and cis- pQTLs. On the one hand, using only cis-pQTLs reduces the risk of violations of the MR assumptions due to pleiotropy, since the pathway from the genetic variant to the protein level is relatively straightforward. On the other hand, for certain proteins, inclusion of trans-pQTL can increase statistical power.

The figure below from Donoghue et al.2 shows the proportion of variance in levels of asthma-relevant proteins explained by cis vs trans pQTL. For certain proteins, it is necessary to use trans pQTLs to explain an adequate proportion of the variance.

pqtl_cis_vs_trans

GWAS Summary Statistics

In addition to tables of pQTLs, the PPP has also released full GWAS summary statistics for each protein measured. There are available at Synapse. These summary statistics are useful for colocalization analysis. For example, after one finds MR evidence suggesting that levels of a protein causally affect a phenotype, one can use colocalization to further support or refute a causal connection3.

Genetics Podcast Interview About PPP

  1. Benjamin B Sun, Joshua Chiou, Matthew Traylor, Christian Benner, Yi-Hsiang Hsu, Tom G Richardson, Praveen Surendran, Anubha Mahajan, Chloe Robins, Steven G Vasquez-Grinnell, and others. Plasma proteomic associations with genetics and health in the UK Biobank. Nature, 622(7982):329–338, 2023. URL: https://www.nature.com/articles/s41586-023-06592-6. ↩↩

  2. Lauren J Donoghue, Christian Benner, Diana Chang, Flaviyan Jerome Irudayanathan, Rion K Pendergrass, Brian L Yaspan, Anubha Mahajan, and Mark I McCarthy. Integration of biobank-scale genetics and plasma proteomics reveals evidence for causal processes in asthma risk and heterogeneity. Cell Genomics, 2025. URL: https://www.cell.com/cell-genomics/fulltext/S2666-979X(25)00096-5. ↩

  3. Verena Zuber, Nastasiya F Grinberg, Dipender Gill, Ichcha Manipur, Eric AW Slob, Ashish Patel, Chris Wallace, and Stephen Burgess. Combining evidence from mendelian randomization and colocalization: Review and comparison of approaches. The American Journal of Human Genetics, 109(5):767–782, 2022. URL: https://www.cell.com/ajhg/fulltext/S0002-9297(22)00149-5. ↩