Mapping quantitative trait loci by an extension of the Haley-Knott regression method using estimating equations

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• The Haley-Knott (HK) regression is a widely used approximation for interval mapping (IM) of quantitative trait loci (QTL) in experimental crosses.
• While computationally efficient, the HK method can be inaccurate, especially with epistasis or linked QTL, and suffers from biased residual variance estimation.

Purpose of the Study:

• To present an extension of the HK method, termed the estimating equation (EE) method, that improves QTL mapping accuracy and precision.
• To evaluate the performance of the EE method across diverse genetic models and data structures.

Main Methods:

• Developed an extension of the HK method using estimating equations based on both means and variances.
• Conducted extensive computer simulations to assess the EE method's performance compared to the HK method.

Main Results:

• The EE method is more efficient than the HK method for normally distributed phenotypes.
• Simulations demonstrated the EE method's robust performance across various genetic models, including nonnormal phenotypes, missing data, epistasis, and linkage.

Conclusions:

• The EE method retains the computational speed and robustness of the HK method.
• The EE method provides a more accurate and precise approximation to IM for QTL detection, outperforming the HK method in parameter estimation and power.