Page Banner

United States Department of Agriculture

Agricultural Research Service

Title: Detection and Parameter Estimation for Quantitative Trait Loci Using Regression Models and Multiple Markers

Authors
item Da, Yang - UNIVERSITY OF MINNESOTA
item Vanraden, Paul
item Schook, Larry - UNIVERSITY OF MINNESOTA

Submitted to: Genetic Selection Evolution
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: April 12, 2000
Publication Date: N/A

Interpretive Summary: Many of the genes that influence important traits are still unknown, but their inheritance can be studied using genetic markers. New measures of the association between markers and the genes of interest (QTL) were developed. For many patterns of markers on a chromosome, the effect of a particular QTL can be determined from just the two markers on either side of the QTL. All other markers can be ignored. Effects of two QTL on the same chromosome can also be estimated independently if at least two markers are located between them. Genetic effects often are treated as random effects, and the effects of individual QTL can also be treated as random. This can be split into QTL variance associated with the markers and the recombination between markers and QTL. The total genetic variance is the sum of the QTL variances associated with markers, the QTL variance not detected because of distance from markers, plus genetic effects on other chromosomes or too small to be detected. These new measures provide a simpler method to estimate effects of genes and are more consistent with previous genetic theory.

Technical Abstract: A strategy of discrete interval mapping based on a minimal conditional regression analysis is proposed to obtain independent testing and parameter estimation for each quantitative trait locus (QTL). This strategy determines whether a chromosome contains one QTL or more than one QTL and whether independent testing and parameter estimation are available for each QTL from current genetic markers in the data set. For QTLs with independent testing and parameter estimation, three sets of closed-form formulations that do not require numerical maximization are derived to estimate the QTL location and effect. The formulations involve two flanking markers if the chromosome contains only one QTL, two flanking markers plus a conditional marker if the chromosome contains at least one additional QTL to one side of the target QTL, and two flanking markers plus two conditional markers if the chromosome contains at least one QTL on either side of the target QTL. Formulations for variance of each QTL are derived, and the size of the QTL effect is twice of the square root of the QTL variance. The total QTL variance is obtained as a sum of individual QTL variances. Simulation data show that these formulations could be a powerful statistical tool for fine QTL mapping. With 1000 observations, a QTL could be mapped to a narrow chromosome region of 1.5 cM if no linked QTL is present, and to a 2.8 cM chromosome region if either side of the target QTL has at least one linked QTL. As sample size increases, a QTL could be mapping chromosome region of less than 1cM.

Last Modified: 10/25/2014
Footer Content Back to Top of Page