Statistical models for estimating the genetic basis of repeated measures and other function-valued traits.

摘要: The genetic analysis of characters that are best considered as functions some independent and continuous variable, such age, can be a complicated matter, simple efficient procedure is desirable. Three methods common in the literature: random regression, orthogonal polynomial approximation, character process models. goals this article (i) to clarify relationships between these methods; (ii) develop general extension model relaxes correlation stationarity, its most stringent assumption; (iii) compare contrast techniques evaluate their performance across range actual simulated data. We find model, described 1999 by Pletcher Geyer, successful method for data examined study. It provides reasonable description wide different covariance structures, it results models Our suggests variance Drosophila mortality declines with while constant at all ages reproductive output. For growth beef cattle, however, increases linearly from birth, correlations high observed ages.