Complex pedigree analysis to detect quantitative trait loci in dairy cattle

M.C.A.M. Bink

Research output: Thesisinternal PhD, WU


In dairy cattle, many quantitative traits of economic importance show phenotypic variation. For breeding purposes the analysis of this phenotypic variation and uncovering the contribution of genetic factors is very important. Usually, the individual gene effects contributing to the quantitative genetic variation can not be distinguished. Developments in molecular genetics, however, have resulted in the identification of polymorphic sites in the genome, which are called genetic markers. Genetic markers have opened the way to follow segregation of chromosomal segments in families. Through the use of these genetically marked chromosomal segments, detection and mapping the genes affecting quantitative traits ("quantitative trait loci" or "QTL") becomes possible. After identifying QTL, genetic markers may, for example, be used to select animals at a younger age and/or to improve the accuracy of predictions of genetic merit.

The aim of this thesis is to contribute to the efficient utilization of genetic marker and quantitative trait data in detecting and utilizing single QTL in complex pedigrees in dairy cattle breeding programs. Implementation of marker-assisted selection in dairy cattle has been hampered by the lack of identified QTL, and the lack of efficient methods for marker-assisted genetic evaluation for situations with incomplete marker data. The development of statistical methods forms the core of this thesis. Methodology is based on Bayes theory and implemented via Markov chain Monte Carlo algorithms, such as the Metropolis-Hastings algorithm and the Gibbs sampler.

Throughout this thesis, a mixed linear model with two random genetic components, i.e., effects due to a marked QTL and residual polygenes, was assumed. These components are assumed to be normally distributed and independent in the base population. To arrive at a flexible method for full pedigree analysis, an animal model is taken as the starting point. Covariances among genetic effects of related individuals are taken into account via the numerator relationship matrix for polygenes and the gametic relationship matrix for QTL.In most chapters, the developed methodology is empirically tested by the use of simulated data. In one chapter, however, experimental data on bovine chromosome six is analyzed to estimate the position and size of a putative QTL for protein percent.

At bovine chromosome six , a QTL for protein percent was identified. The most likely position of this QTL is similar to that previously reported by Spelman and co-workers. The presence of a second putative QTL for protein percent is uncertain and requires further research probably with a two-QTL model. In the general discussion, the presented Bayesian method is compared to other methods for QTL analysis in complex pedigrees. Our method at this moment is unique in being able to handle complex pedigrees in outbred populations with missing marker data. The general discussion is completed with a brief review of issues related to practical implications for marker-assisted genetic evaluation in dairy cattle breeding schemes.

Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Brascamp, E.W., Promotor, External person
  • Quaas, R.L., Promotor, External person
  • van Arendonk, J.A.M., Promotor
Award date4 Sep 1998
Place of PublicationS.l.
Print ISBNs9789054858973
Publication statusPublished - 1998


  • dairy cattle
  • pedigree
  • breeding value
  • selective breeding
  • genetic variation
  • genetic markers
  • quantitative traits
  • methodology
  • algorithms
  • monte carlo method


Dive into the research topics of 'Complex pedigree analysis to detect quantitative trait loci in dairy cattle'. Together they form a unique fingerprint.

Cite this