Fig. 1: Overview of study design.
From: A survey of pathogenic involvement in non-communicable human diseases
The UK Biobank (UKB) was used as the discovery cohort (left). 426 diseases with sufficient sample counts in UKB (along with positive and negative controls) were tested for association with the 45 antibody titers representing immune responses to 20 unique pathogens across 9429 UKB participants. Models were adjusted for any of ten additional health-related and sociodemographic variables that were determined to be confounding for a particular antibody-disease pair. The antibody-disease pair results were collapsed to the pathogen-disease level by selecting the most significant antibody-disease result to represent that pathogenʼs association with the disease. Significant pathogen-disease pairs identified in UKB were tested in the independent TriNetX (TNX) cohort (right). Pathogen-disease pairs that were significant in both the discovery and the replication cohort were considered replicated pairs.
