The Additive Genetic Gamma Frailty Model

From MaRDI portal





The additive genetic gamma frailties (genetic frailties) \(z_1,\dots,z_n\) for \(n\) individuals are defined as linear combinations of independent gamma distributed random variables with the properties \(E(z_i)=1\), \(i=1,\dots,n\), \(var(z_1,\dots,z_n)=A\sigma^2_z\), where \(A\) is the numerator relationship matrix. Two examples which illustrate the idea of construction of genetic frailties are presented. A general algorithm for construction of additive genetic gamma frailties in any pedigree of \(n\) individuals is given in an appendix. The EM algorithm can be used to estimate the parameters of the model. Calculations are similar to the EM algorithm in shared frailty models but, however, the E-step is not correspondingly simple. Data from the ``Danish adoptive register on triplets (father, mother, child) are analysed using the genetic frailty model as well as bivariate relationships (father, child) and (mother, child). The results are compared to those obtained with the help of the shared frailty model developed by \textit{G. Nielsen, R. Gill, P. Andersen} and \textit{T. Sorensen} [Scand. J. Stat. 19, No. 1, 25-43 (1992; Zbl 0747.62093)]. Advantages of the genetic gamma frailty model are outlined.











This page was built for publication: The Additive Genetic Gamma Frailty Model

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4231301)