Fine mapping of multiple QTL using combined linkage and linkage disequilibrium mapping – A comparison of single QTL and multi QTL methods

Abstract

-Two previously described QTL mapping methods, which combine linkage analysis (LA) and linkage disequilibrium analysis (LD), were compared for their ability to detect and map multiple QTL. The methods were tested on five different simulated data sets in which the exact QTL positions were known. Every simulated data set contained two QTL, but the distances between these QTL were varied from 15 to 150 cM. The results show that the single QTL mapping method (LDLA) gave good results as long as the distance between the QTL was large (> 90 cM). When the distance between the QTL was reduced, the single QTL method had problems positioning the two QTL and tended to position only one QTL, i.e. a "ghost" QTL, in between the two real QTL positions. The multi QTL mapping method (MP-LDLA) gave good results for all evaluated distances between the QTL. For the large distances between the QTL (> 90 cM) the single QTL method more often positioned the QTL in the correct marker bracket, but considering the broader likelihood peaks of the single point method it could be argued that the multi QTL method was more precise. Since the distances were reduced the multi QTL method was clearly more accurate than the single QTL method. The two methods combine well, and together provide a good tool to position single or multiple QTL in practical situations, where the number of QTL and their positions are unknown. fine mapping / multiple QTL / simulations

Effects of the number of markers per haplotype and clustering of haplotypes on the accuracy of QTL mapping and prediction of genomic breeding values

“…Since the data varied from replicate to replicate in terms of the distance between and the size of QTL, a proper investigation of the ability to separate nearby QTL was not possible based on our results. However, Uleberg and Meuwissen [22] have shown that a multiple QTL model comparable to the model used in our study could distinguish two QTL located 15 cM apart.…”

Genome-wide association study to identify chromosomal regions associated with antibody response to Mycobacterium avium subspecies paratuberculosis in milk of Dutch Holstein-Friesians

“…This increase in power is similar to the principles of multiple QTL mapping (Jansen, 1993) and can lead to the detection of new chromosomal regions associated with the trait of interest, which might not be detected in a single-SNP analysis and may explain differences found between studies. Additionally, 2 simulation studies (Sillanpää and Arjas, 1998;Uleberg and Meuwissen, 2007) showed that the confidence interval became smaller when information from all QTL positions was used in multiple-QTL analysis compared with single-QTL analysis.…”

Makeup of the genetic correlation between milk production traits using genome-wide single nucleotide polymorphism information

“…About 0.5 to 1.0% of the SNP had a significant effect on one or more traits, ranging from 200, 308, and 334 SNP for protein, fat, and milk yield, and 318 and 387 for fat and protein percentage, respectively. Pryce et al ( 2010) reported a remarkably similar number (Uleberg and Meuwissen, 2007;Veerkamp et al, 2010;Calus et al, 2011b).…”