Genetic control of the human brain proteome

We generated an online brain pQTL resource for 7,376 proteins through the analysis of genetic and proteomic data derived from postmortem samples of the dorsolateral prefrontal cortex of 330 older adults. The identified pQTLs tend to be non-synonymous variation, are over-represented among variants associated with brain diseases, and replicate well (77%) in an independent brain dataset. Comparison to a large study of brain eQTLs revealed that about 75% of pQTLs are also eQTLs. In contrast, about 40% of eQTLs were identified as pQTLs. These results are consistent with lower pQTL mapping power and greater evolutionary constraint on protein abundance. The latter is additionally supported by observations of pQTLs with large effects' tending to be rare, deleterious, and associated with proteins that have evidence for fewer protein-protein interactions. Mediation analyses using matched transcriptomic and proteomic data provided additional evidence that pQTL effects are often, but not always, mediated by mRNA. Specifically, we identified roughly 1.6 times more mRNA-mediated pQTLs than mRNA-independent pQTLs (550 versus 341). Our pQTL resource provides insight into the functional consequences of genetic variation in the human brain and a basis for novel investigations of genetics and disease.

Automated summary

The study generated an online brain pQTL resource for 7,376 proteins by analyzing genetic and proteomic data from the dorsolateral prefrontal cortex of 330 older adults. The identified pQTLs tend to be non-synonymous variations, over-represented among brain disease-associated variants, and show good replication in an independent brain dataset. Comparison with brain eQTLs revealed that most pQTLs are also eQTLs, suggesting lower pQTL mapping power and greater evolutionary constraint on protein abundance.