Genome-wide mediation analysis: bridging the dividebetween genotype and phenotype via transcriptomicdata in maize

Mapping genotype to phenotype is an essential topic in genetics and genomics research. As the Omics data become increasingly available, genome-wide association study (GWAS) has been widely applied to establish the relationship between genotype and phenotype. However, signals detected by GWAS usually span broad genomic regions with many underneath candidate genes, making it challenging to interpret and validate the molecular functions of the candidate genes. Under the context of genetics research, we hypothesized a causal chain from genotype to phenotype partially mediated by intermediate molecular processes, i.e., gene expression. To test this hypothesis, we applied the high dimensional mediation analysis, a class of causal inference method with an assumed causal chain from the exposure to the mediator to the outcome, and implemented it to the maize diversity panel (N=280 lines). Using 40 publicly available agronomic traits, 66 newly generated metabolic traits, and published RNA-seq data from seven different tissues, we detected N=736 unique mediating genes, explaining an average of 12.7% phenotypic variance due to mediation. Noticeably, 83/736 (11%) genes were identified in mediating more than one trait, suggesting the prevalence of pleiotropic mediating effects. Among those pleiotropic mediators, benzox-azinone synthesis 13 (Bx13), a well-characterized gene encoding a 2-oxoglutarate-dependent dioxygenase, was identified mediating 40 agronomic and metabolic traits in different tissues. Further genetic and genomic analyses of the Bx13 and adjacent mediating genes suggested a 3D co-regulation modulation likely affect their expression levels and eventually lead to phenotypic consequences. Our results suggested the genome-wide mediation analysis is a powerful tool to integrate Omics data in providing causal inference to connect genotype to phenotype.