Progress 09/01/03 to 08/31/07
Outputs
Outputs: Proposal objectives were: 1. Extend pedigree-based QTL methods to account for changes in allelic effects resulting from epistatic interactions, mutation or recombination. 2. Evaluate the power to detect and distinguish among mechanisms that produce variance and optimize selection design for mechanism detection and utilization. 3. Interact with potential user groups to determine useful functionality for applied and research purposes. Of these objectives, only parts of objectives 1 and 2 were accomplished. In particular: 1. Theory was developed and implemented to detect epistasis in the form of QTL x genetic background interaction in complex pedigrees, and the power and accuracy of estimation were evaluated in random mating populations and in pedigrees of selfing species. 2. Theory has been developed and implemented to detect mutation and / or recombination between tightly linked causal polymorphisms in complex pedigrees. The method has not been applied to real or simulated data as of now. Because of hiring problems detailed in Project Modifications, we ran out of time before we could address Objective 3.
Impacts
Identifying genetic mechanisms that generate genetic variance upon which selection can act will lead to greater understanding of longterm selection responses. This greater understanding will enable plant and animal breeders to develop breeding schemes that more effectively account for those mechanisms and therefore produce greater selection response. Identifying QTL allele interactions with the genetic background will improve our ability to target specific loci for transfer to specific genetic backgrounds. It will improve our ability to predict the outcomes of marker assisted backcrossing efforts. Understanding the frequency and effect size of mutations in a breeding program would be a very important step toward optimising long-term selection approaches to take advantage of mutations. Such understanding would also have a large impact on our understanding of evolution through natural selection. The impacts are at this point primarily theoretical. We need empirical
application to know how prevalent these mechanisms are, and therefore how valuable are methods to detect them.
Publications
- Jannink, J.L. 2007. Identifying Quantitative Trait Locus by Genetic Background Interactions in Association Studies. Genetics 176:553-561.
- Jannink, J.-L. 2007. QTL by genetic background interaction: predicting inbred progeny value. Euphytica Accepted: 11 July 2007.
- Zhong, S., and J.-L. Jannink. 2007. Using quantitative trait locus results to discriminate among crosses based on their progeny mean and variance. Genetics 177:567-576.
|
Progress 01/01/06 to 12/31/06
Outputs
We completed the theory for the identification of quantitative trait locus (QTL) interaction with genetic background for biallelic models. We tied this theory back to more traditional quantitative genetic models of epistasis (interactions between genetic loci that affect quantitative traits). We developed software to implement the theory and estimated the strength of the effects of QTL by genetic background interaction. We performed simulations to document the power of the method to detect interactions. In these same simulations, we determined whether the model was better able to predict the genotypic value of progeny than a standard additive model. In some cases it was, although simulations showed that the additive model performs well even when genetic interactions do occur. We completed work on theory for the construction of gametic relationship matrices that account for mutation in a complex pedigree and implemented that theory in software. We have performed
simulations to determine if this random-model approach to identifying mutations can work. The assessment will come in the form of a likelihood ratio test, but it will not be a standard test because both mutated and non-mutated models have the same number of degrees of freedom. We are currently analyzing simulation results to determine our prospects.
Impacts
This program plays a major role in assembling barley genomic information into one database. The shear amount of data requires 'next-generation' databases to efficiently manage information so plant breeders can make improvements in disease and stress resistance, food and nutrition. This research emphasized data analysis and educational aspects of the project. Improvements in barley varieties will come from accurate identification of which varieties have the best genetics and which crosses between varieties will produce the best combinations. New analyses that combine field and malting observations with information on the barley genes and variety origins from The Hordeum Toolbox will help breeders decide which varieties to keep and which to cross. The coordinated project addressed a pressing need for barley -- identifying the genes that control economically important traits in breeding material and enabling breeders to use them to bring better barley to farmers and
consumers. All results of the research will be available to scientists, growers and the food industry through the project's public Web site at http://www.barleycap.org. Understanding the frequency and effect size of mutations in a breeding program will be an important step toward optimizing long-term selection approaches to take advantage of mutations. Such understanding also will have a large impact on the understanding of evolution through natural selection.
Publications
- Verhoeven, K.J.F., Jannink, J.-L. and McIntyre, L.M. 2006. Using mating designs to uncover QTL and the genetic architecture of complex traits. Heredity. 96:139-149.
|
Progress 01/01/05 to 12/31/05
Outputs
As presented earlier, theory for the identification of QTL interaction with the genetic background was incomplete. In this past year we have completed this theory for biallelic models. We are in the process of constructing algorithms to estimate mixed models accounting for a random effect that is the deviation of the effect of an associated QTL allele due to genetic background. We have developed new theory for the construction of gametic relationship matrices that account for the possibility of mutation in the pedigree. The theory requires the estimation of segregation indicators, that is, the probabilities that progeny received either grand-maternal or grand-paternal alleles from each parent. We have developed software to estimate these segregation indicators. We are in the process of constructing the algorithms to estimate mixed models using the gametic relationship matrix.
Impacts
Identifying genetic mechanisms that generate genetic variance upon which selection can act will lead to greater understanding of longterm selection responses. This greater understanding will enable plant and animal breeders to develop breeding schemes that more effectively account for those mechanisms and therefore produce greater selection response. Identifying QTL allele interactions with the genetic background will improve our ability to target specific loci for transfer to specific genetic backgrounds. It will improve our ability to predict the outcomes of marker assisted backcrossing efforts.
Publications
- Jannink, J.-L. 2005. Selective phenotyping to accurately map QTL. Crop Sci. 45:901-908.
|
Progress 01/01/04 to 12/31/04
Outputs
We have further developed statistical theory for the detection of marker by genetic background interaction using a relationship matrix that shows the relatedness of all individuals in the pedigree. We have developed software to perform the calculations required by the theory. Simulation studies showing the power of this approach to detect and assess the effects of genes that interact with the genetic background will be underway in the next month.
Impacts
The research proposed here will enable us to observe the functioning of mechanisms that generate genetic variance within populations under selection. This research is significant because application of the resultant knowledge is expected to lead to new plant breeding strategies capable of maximizing gains from selection by enhancing the generation of genetic variance. We will develop new methods capable of accomplishing marker data analysis in pedigreed populations under selection. These methods are expected to increase the impact of DNA-marker technologies on plant improvement programs.
Publications
- Jannink, J.-L. and Fernando, R.L. 2004. On the Metropolis-Hastings acceptance probability to add or drop a QTL in MCMC-based Bayesian analyses. Genetics 166:641-643.
|
Progress 01/01/03 to 12/31/03
Outputs
This project was only funded for 3 months in 2003 and we have little progress to report. In the process of going over theory in preparation for software development, we discovered a mistake in previous Bayesian QTL analysis methods presented in the literature and correcting this mistake has been the subject of a publication.
Impacts
The manuscript to be published in 2004 will improve inferences on the number of QTL segregating in a population, thereby leading to improved marker assisted selection programs, and enhanced crop performance.
Publications
- No publications reported this period
|
|