Genome-Wide Approach to Measure Variant-Based Heritability of Drug Outcome Phenotypes

Area of Science:

• Genetics
• Pharmacology

Background:

• Pharmacogenomic studies often focus on variants with large effect sizes in drug-related genes.
• These large-effect variants explain only a small portion of the genetic variability in drug response phenotypes.

Purpose of the Study:

• To quantify the narrow-sense heritability ($$) of 12 drug outcome phenotypes using genome-wide common variation.
• To determine the contribution of small, moderate, and large effect variants to the heritability of drug response phenotypes.

Main Methods:

• Utilized a Bayesian hierarchical mixed model (BayesR) to analyze genome-wide common variants in individuals of European ancestry.
• Modeled the distribution of variant effect sizes for pharmacodynamic and pharmacokinetic phenotypes across cardiovascular drugs, antibiotics, and immunosuppressants.
• Partitioned heritability ($$) into contributions from variants with different effect sizes.

Main Results:

• Heritability estimates ($$) for the 12 phenotypes ranged from 0.05 to 0.59.
• Small-effect and moderate-effect variants collectively accounted for the majority (61-95%) of the heritability for all studied drug phenotypes.
• A median of 3,347 variants contributed to the heritability of each phenotype.

Conclusions:

• Drug outcome phenotypes are highly polygenic, meaning they are influenced by numerous genes with small effects.
• Larger genome-wide association studies are necessary to identify more variants and improve the clinical prediction of drug outcomes using genomic data.