Fig. 4: Putative causal relationships and phenotypic associations between DMR-correlated proteins and asthma-related outcomes.

Evidence for the putative causal relationship was obtained using two-sample mendelian randomization (MR Egger, weighted median, inverse variance weighted (IVW), weighted mode) with pQTLs from deCODE as genetic instruments. Phenotypic associations were obtained using linear regression. The analyses to identify phenotypic association in the MARC-35 cohort were conducted in a small subset of participants with available data (recurrent wheezing: 57 cases, 50 non-cases; asthma: 50 cases, 59 non-cases).