Summary statistics for pQTL in UKB-PPP

Based on strict quality control, we only included 33,325 European participants (17,988 females and 15,337 males), with 2,919 plasma proteins imputed using the mean value. We introduce ProteoNexus (https://www.proteonexus.com/), a new platform that characterizes genetic architecture, constructs the prediction model for incident diseases, and estimates the mediation effect of plasma proteins using UK Biobank (application ID: 144904). In pQTL analysis, ProteoNexus used GEMMA to estimate the summary statistics by adjusting covariables (i.e., age, sex, and the top 18 genetic principal components) and employed SuSiE to calculate probability in posterior (PIP) for each SNP and to estimate the credible sets for each locus.