Identifying Multi-Omics Causers and Causal Pathways for Complex Traits.

Area of Science:

• Molecular Biology
• Genetics
• Systems Biology

Background:

• The central dogma describes DNA → RNA → protein → trait.
• Genome-wide association studies (GWAS) identify genetic variants linked to traits but ignore intermediate omics layers.
• Conventional association studies may lack power and identify non-causal variants.

Purpose of the Study:

• To model the central dogma using a mediate causal model.
• To analytically and numerically demonstrate the power of different omics layers in trait association.
• To highlight the necessity of integrating multi-omics data for causal variant discovery.

Main Methods:

• Developed a mediate causal model based on the central dogma.
• Analytically proved the relationship between omics level distance and correlation magnitude.
• Numerically demonstrated power differences using random and extreme sampling schemes.

Main Results:

• The correlation magnitude decreases as the omics level is more remote from the physiological trait.
• Proteome-trait correlation tests showed higher power than transcriptome-trait tests.
• Transcriptome-trait tests were more powerful than genotype-trait association tests.

Conclusions:

• Integrating DNA, RNA, and protein expression data with causal inference is crucial.
• This multi-omics approach is necessary for a comprehensive understanding of genetic contributions to phenotype variation.
• Proteomics offers a more powerful layer than transcriptomics or genomics for trait association studies.