A comparison of interval mapping procedures including the QTL-cluster mapping

Area of Science:

• Quantitative Genetics
• Genomic Analysis
• Statistical Genetics

Background:

• Accurate quantitative trait loci (QTL) mapping is crucial for understanding genetic architecture.
• Interval mapping (IM) and regression interval mapping (RIM) are common methods, but can face challenges with linked QTLs.
• QTL-cluster mapping (CM) offers an alternative approach considering linked QTLs within marker intervals.

Purpose of the Study:

• To investigate and compare the performance of IM, RIM, and CM in estimating QTL position and effect.
• To analyze the behavior of log-likelihood surfaces under various genetic models and linkage phases.
• To determine the most robust method for QTL analysis, particularly for marker-assisted selection.

Main Methods:

• Deterministic sampling was employed to analyze log-likelihood surfaces.
• Simulations were conducted using a half-sib design under diverse genetic models.
• The study compared IM, RIM, and CM, focusing on scenarios with single or closely linked QTLs within marker intervals.

Main Results:

• IM and RIM showed clear log-likelihood maxima and unbiased QTL effect estimates for single QTLs or coupled linked QTLs.
• For repulsion-phase linked QTLs, IM and RIM produced a log-likelihood ridge, indicating dependency and biased estimates.
• QTL-cluster mapping (CM) consistently provided unbiased QTL effect estimates across all tested genetic models.

Conclusions:

• CM is a more robust method for estimating QTL effects at the segment level, especially when dealing with linked QTLs in various phases.
• IM and RIM can be unreliable for linked QTLs in repulsion phase, potentially leading to biased estimates.
• CM is recommended for marker-assisted selection applications over individual closely linked QTL mapping.