Page Banner

United States Department of Agriculture

Agricultural Research Service

Title: Parameter Estimation of Quantitative Trait Loci Using Regression Models Anddiscrete Interval Mapping

Authors
item Da, Yang - UNIV OF MINNESOTA
item Vanraden, Paul
item Beattie, Craig - UNIV OF MINNESOTA
item Schook, Lawrence - UNIV OF MINNESOTA

Submitted to: International Congress of Genetics Proceedings
Publication Type: Abstract Only
Publication Acceptance Date: June 1, 1998
Publication Date: N/A

Technical Abstract: Distance between a marker and a quantitative trait locus (QTL) and the QTL effect are important parameters in mapping QTL. An accurate estimate of QTL-marker distance is crucial for finding the physical location of a QTL. A new approach for parameter estimation was developed using regression models and discrete interval mapping. Genotypic value of a QTL is partitioned by genetic makers into marker effect and recombination residual; variance (Var) of QTL genotypic value is partitioned into Var(marker effect) and Var(recombination residual). Based on these partitions, marker-QTL recombination frequency can be expressed as .5 times [1 minus [square root of [Var(marker effect) divided by [Var(marker effect) + Var(recombination residual)]]]]. However, Var(recombination residual) is confounded with Var(random residual). A strategy was developed to recover Var(recombination residual) from Var(phenotypic residual) by expressing Var(recombination residual) in terms of Var(marker) from two single-marker analyses and one two-marker analysis. Estimates of these Var(marker) can be obtained from marker regression coefficients or variance components. Once Var(recombination residual) is recovered, Var(QTL) is available, and QTL effect is simply 2 times Var(QTL genotypic value). For multiple QTL, a strategy of discrete interval mapping was developed to obtain independent estimates of QTL parameters for multiple-linked QTL using mathematical formulations for individual QTL. This strategy also provides estimates of total Var(QTL) and heritability using QTL information rather than the traditional approach of resemblance between relatives and opens a new window of opportunity to obtain an exhaustive description of quantitative variations.

Last Modified: 9/1/2014
Footer Content Back to Top of Page