Unbalanced Genetic Data Analyses: Genetic Model Evaluations and Applications.

Fully utilizing genetic information is important for crop improvement and relies on appropriate data analyses; however, genetic data structures collected from different breeding programs may often be unbalanced due to a large of number parents employed, constraints of seed supply, and/or unpredictable factors. Traditional methods are challenged by such unbalanced data structures and thus valuable genetic information may be under-utilized. Therefore, it is valuable to assess genetic analyses for unbalanced data structures by utilizing new methodologies. Using mixed linear model approaches, several commonly used genetic models, including additive-dominance model and their extensions regarding their statistical properties were investigated by Monte Carlo simulations. As for demonstration, actual unbalanced breeding data will be analyzed. Detailed results from model evaluations and actual data will be presented.