Complex genetic effects in quantitative trait locus identification: a computationally tractable random model for use...

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Traditional quantitative trait loci (QTL) mapping often treats QTL as fixed effects, differing from random genetic effect models.
• Existing computationally intensive methods allow QTL to be treated as random for additive and dominance effects.

Purpose of the Study:

• To extend variance component methods (VCM) for mapping multiple QTL.
• To develop a computationally efficient approach for estimating genetic effects and covariances in F(2) populations.

Main Methods:

• Extended VCM to map multiple QTL in an F(2) crossbred population.
• Estimated individual effects and marker-derived genetic covariances.
• Approximated genetic covariances by replacing individual genetic effects with average effects for marker classes to reduce computational burden.

Main Results:

• The reduced model significantly increased the speed of variance component estimation and residual log-likelihood evaluation.
• Simulations showed the reduced model is competitive with existing methods in QTL position estimation and experimental power.
• The method effectively handles pairwise epistatic effects.

Conclusions:

• The developed VCM extension offers a computationally efficient solution for multiple QTL mapping.
• Approximating genetic covariances substantially reduces computational demands without compromising accuracy.
• This approach enhances the feasibility of complex genetic analyses involving multiple QTL and epistasis.