Bayesian genomic-enabled prediction as an inverse problem

Genomic-enabled prediction in plant and animal breeding has become an active area of research. Many prediction models address the collinearity that arises when the number (p) of molecular markers (e.g. single-nucleotide polymorphisms) is larger than the sample size (n). Here we propose four Bayesian...

Descripción completa

Detalles Bibliográficos
Autores: Cuevas, J., Perez-Elizalde, S., Soberanis, V., Perez-Rodriguez, P., Gianola, D., Crossa, J.
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2014
País:México
Institución:Centro Internacional de Mejoramiento de Maíz y Trigo
Repositorio:Repositorio Institucional de Publicaciones Multimedia del CIMMYT
OAI Identifier:oai:repository.cimmyt.org:10883/34648
Acceso en línea:https://hdl.handle.net/10883/34648
Access Level:acceso abierto
Palabra clave:AGRICULTURAL SCIENCES AND BIOTECHNOLOGY
Bayesian Regression
Prior Distribution
GenPred
Shared Data Resources
Genomic Selection
BAYESIAN THEORY
SHRINKAGE
REGRESSION ANALYSIS
DATA
MARKER-ASSISTED SELECTION
Genetic Resources
Descripción
Sumario:Genomic-enabled prediction in plant and animal breeding has become an active area of research. Many prediction models address the collinearity that arises when the number (p) of molecular markers (e.g. single-nucleotide polymorphisms) is larger than the sample size (n). Here we propose four Bayesian approaches to the problem based on commonly used data reduction methods. Specifically, we use a Gaussian linear model for an orthogonal transformation of both the observed data and the matrix of molecular markers. Because shrinkage of estimates is affected by the prior variance of transformed effects, we propose four structures of the prior variance as a way of potentially increasing the prediction accuracy of the models fitted. To evaluate our methods, maize and wheat data previously used with standard Bayesian regression models were employed for measuring prediction accuracy using the proposed models. Results indicate that, for the maize and wheat data sets, our Bayesian models yielded, on average, a prediction accuracy that is 3% greater than that of standard Bayesian regression models, with less computational effort.