|Zhang, Z. - CORNELL UNIV., ITHACA|
|Todhunter, R. - CORNELL UNIV., ITHACA|
|Van Vleck, Lloyd|
Submitted to: Journal of Animal Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: November 1, 2006
Publication Date: March 1, 2007
Citation: Zhang, Z., Todhunter, R.J., Buckler Iv, E.S., Van Vleck, L.D. 2007. Technical note: Use of marker-based relationships with multiple-trait derivative-free restricted maximal likelihood. Journal of Animal Science. 85:881-885. Interpretive Summary: As advances in genomic research lead to increased availability and affordability of genotyping, relationships based on molecular markers are becoming more obtainable in both animal and plant breeding programs. These marker-based relationships can be superior to conventional pedigree-based relationships in controlling for false positives in QTL association studies. In lieu of probabilities based on identity by descent, marker-based kinships may provide a better alternative in situations where pedigrees are not deep, are not complete (e.g., father or mother is unknown) or are unknown. To exploit these sources of genetic information, algorithms have been developed to estimate kinships based on genetic markers. Several software packages have also been developed. Although the MTDFREML set of programs has been widely used to estimate variance components, to predict breeding values and to study associations between markers and QTL, the initial version of MTDFREML was generally limited to the use of relationships based on pedigree only. One notable exception was study that successfully employed the MTDFREML programs to compare models with line effects treated as either random or fixed factors by using coefficients of coancestry (one-half additive relationships) among 58 maize inbred lines to calculate inverse elements of the relationship matrix. Modification of the set of MTDFREML programs to include an alternative version of the first component will now allow users to either calculate relationships based on standard pedigree files or to use relationships derived directly from molecular markers. This technical note describes the necessary modifications and includes an empirical example of their implementation in a QTL association study of canine hip dysplasia. The implications are that the options of 1) use of an arbitrary relationship matrix derived from markers in common or from portions of a full relationship matrix or 2) use of a traditional set of rules for obtaining the inverse of the augmented relationship matrix will now allow for flexibility in accounting for polygenic effects for analyses of genomic data or for estimating polygenic breeding values and components of variance due to polygenic effects.
Technical Abstract: The widespread use of the set of MTDFREML computer programs for genetic evaluation and estimation of genetic parameters has led to significant improvement in traits of economic importance. The initial version of this software package, however, was generally limited to use of pedigree-based relationships. With continued advances in genomic research and the increased availability of genotyping, relationships based on molecular markers are both obtainable and desirable. The addition of a new component to the MTDFREML set of programs is described which allows users the flexibility to calculate relationships using either standard pedigree files or an arbitrary relationship matrix based on genetic marker information. The strategy behind this modification and its design are described. An application is illustrated in a QTL association study for canine hip dysplasia.