Principled measures and estimates of trait polygenicity

Area of Science:

• Quantitative genetics
• Statistical genetics
• Human genomics

Background:

• The term "polygenicity" is frequently used in genetic studies but lacks a clear, standardized definition.
• Existing quantitative and statistical genetic studies quantify polygenicity, but its precise meaning remains ambiguous.

Purpose of the Study:

• To propose a principled definition of polygenicity encompassing various measures.
• To establish mathematical properties for valid polygenicity measures.
• To estimate polygenicity for 36 human traits and investigate the distribution of these values.

Main Methods:

• Developed a principled definition for polygenicity.
• Identified and validated mathematical properties for polygenicity measures.
• Utilized Fourier mixture regression to estimate three polygenicity measures from genome-wide association studies (GWAS) summary statistics.
• Applied these methods to 36 human traits.

Main Results:

• Proposed a definition of polygenicity with sensible mathematical properties.
• Successfully estimated polygenicity measures for 36 human traits.
• Observed a notable absence of traits with polygenicity values intermediate between Mendelian and highly polygenic extremes.
• Identified a significant gap in the distribution of polygenicity across human traits.

Conclusions:

• The proposed definition provides a rigorous framework for understanding and measuring polygenicity.
• The observed gap in polygenicity distribution suggests potential insights into underlying evolutionary and cellular mechanisms.
• Further research is warranted to explore the generality of this finding and its biological implications.