A comparison of regression interval mapping and multiple interval mapping for linked QTL.

Area of Science:

• Quantitative genetics
• Statistical genomics
• Bioinformatics

Background:

• Mapping quantitative trait loci (QTL) is crucial for understanding complex traits.
• Inbred-line crosses are commonly used for genetic mapping studies.
• Comparing different mapping methods is essential for optimizing experimental design.

Purpose of the Study:

• To compare the performance of regression interval mapping and multiple interval mapping for linked QTL.
• To evaluate the power and accuracy of QTL identification and parameter estimation.
• To assess the impact of population size and marker interval on mapping accuracy.

Main Methods:

• Simulation study using genetic models with two linked QTL.
• Data simulation for F(2) populations of varying sizes.
• Comparison based on QTL identification power, position/effect accuracy, and variance estimation.

Main Results:

• Multiple interval mapping demonstrated higher power and more accurate QTL position and effect estimates than regression interval mapping.
• The superiority of multiple interval mapping increased with wider marker intervals and larger population sizes.
• Reducing marker interval size from 10 to 5 cM significantly improved power and estimates for both methods.

Conclusions:

• Multiple interval mapping is a more powerful and accurate method for mapping linked QTL in inbred-line crosses.
• Denser marker intervals (5 cM) are beneficial for improving QTL mapping accuracy, contrary to previous findings for single QTL.
• Standard statistical theory underestimates standard errors for both mapping methods.