Title :
Graphical model representation of pedigree based mixed model
Author_Institution :
Dept. of Animal Sci., Univ. of Ljubljana, Domžale, Slovenia
Abstract :
Pedigree based mixed model represents a simplistic yet robust and powerful model frequently used in animal and plant breeding, evolutionary biology, and human genetics. In the Bayesian setting inference of all model parameters can be performed with the use of well known McMC methods. Algorithms are commonly formulated with matrices, which provides a generic view, but hinders interpretation. Here, a generic graphical model representation is developed. This eases the interpretation of the model and used algorithms. In addition, graphical model formulation provides a way to fit pedigree based mixed model in standard graphical model programs, such as BUGS.
Keywords :
Markov processes; Monte Carlo methods; belief networks; biology computing; genetic engineering; inference mechanisms; matrix algebra; solid modelling; BUGS programs; Bayesian setting inference; Markov chain Monte Carlo methods; animal breeding; evolutionary biology; graphical model representation; human genetics; matrices; pedigree based mixed model; plant breeding; Additives; Animals; Bayesian methods; Biological system modeling; Genetics; Graphical models; Markov processes; McMC; animal model; graphical model; pedigree;
Conference_Titel :
Information Technology Interfaces (ITI), 2010 32nd International Conference on
Conference_Location :
Cavtat/Dubrovnik
Print_ISBN :
978-1-4244-5732-8