Resumen

BANDERA-FERNANDEZ, Evelyn  y  PEREZ-PELEA, Leneidy. Generalized linear mixed models. Its application in plant breeding. cultrop [online]. 2018, vol.39, n.1, pp. 127-133. ISSN 1819-4087.

ABSTRACT Frequently, in agricultural research, experimental data do not satisfy the assumptions of general linear models, making the recommended transformations very few effectivel. Generalized Linear Mixed Models (GLMMs) provide a way of analysis for those data that are correlated and that does not necessarily require that the evaluated variable be normally distributed and but to a distribution of the exponencial family (Gamma, Poisson, Binomial among others). The objective of this review was to present the applications of generalized linear mixed models in breeding programs. This model has been focused fundamentally in three directions in plant breeding programs: in the prediction of family behavior, in the estimation of variance components and in multi-environment trials. GLMM estimation procedures make it possible to reduce biases when data is incomplete, unbalanced or adjust scattered data and also allow modeling the structure of errors in data from longitudinal measurements. There are several commercial and free programs such as: the GLIMMIX and GENMOD procedure of the SAS package and the Ime4 of the R package that enable the use of The Generalized Linear Mixed Models for most current applications in plant genetics

Palabras clave : plant breeding; quantitative genetics.

        · resumen en Español     · texto en Español     · Español ( pdf )