A structural model of genome-wide association studies

TL;DR

This paper develops a mathematical structural model for genome-wide association studies, linking genome structure to genetic correlation, and validates it with empirical data from mouse crosses.

Contribution

It introduces a novel structural genetic model that explains genome map distances and genetic correlations, validated against empirical mouse data.

Findings

Model explains Haldane map distance as a consequence of genome structure

Close agreement between model predictions and empirical data

Identifies regions for further genetic investigation

Abstract

A structural genetic model incorporating a modern understanding of the genome and common practice in genome-wide association studies is derived mathematically. The model shows the Haldane map distance as a direct consequence of the structure of the genome. An expression for genetic correlation is derived under the model and compared to data resulting from the BSB mouse cross. A correlation test plot is introduced for this comparison and shows the close agreement of the model and empirical results. Noteworthy departures in this plot indicate regions which warrant further investigation.