Testing quantitative trait locus effects in genetic backcross studies with double recombination occurring

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Quantitative trait locus (QTL) mapping is crucial for understanding genetic contributions to complex traits.
• Traditional methods often assume equal scale distributions, which may not hold true in real genetic data.
• Testing QTL effects in both location and scale is essential for comprehensive genetic analysis.

Purpose of the Study:

• To develop a robust statistical test for detecting quantitative trait locus (QTL) effects in genetic mapping studies.
• To address the unbounded log-likelihood issue in location-scale distribution families without equal scale assumptions.
• To propose a penalized likelihood ratio test (PLRT) as a valid alternative to traditional likelihood ratio tests.

Main Methods:

• Phenotype distributions were modeled using a location-scale distribution family.
• A penalized likelihood ratio test (PLRT) was developed to overcome unbounded log-likelihood problems.
• Null limiting distributions for PLRT and likelihood ratio tests (LRT) were derived.
• Asymptotic distributions under local alternatives were investigated.

Main Results:

• The proposed penalized likelihood ratio test (PLRT) is shown to be a valid method for testing QTL effects.
• The null limiting distribution of the PLRT is a supremum of a chi-square process.
• The study also characterized the null limiting distribution of the LRT under equal scale assumptions.
• Simulation studies confirmed the accuracy of the asymptotic results.

Conclusions:

• The penalized likelihood ratio test (PLRT) provides a reliable approach for detecting quantitative trait locus (QTL) effects in genetic studies, especially when scale parameters vary.
• The derived limiting distributions offer theoretical support for the proposed statistical methods.
• The findings enhance the toolkit for genetic analysis and QTL mapping.