期刊
JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS
卷 24, 期 3, 页码 756-770出版社
AMER STATISTICAL ASSOC
DOI: 10.1080/10618600.2014.948180
关键词
Smoothing; Genetic relationship; Functional principal components; Sparse functional data
资金
- Natural Sciences and Engineering Research Council of Canada (NSERC)
- Meat and Livestock Australia [B.BFG.0050]
We propose a new version of functional data model for analyzing familial related individuals, where the within-subject correlation depends smoothly on a covariate such as age and the between-subject correlation follows family-wise genetic association. Our motivating example concerns measurements of weight as a function of age in sibling cows from independent families. Observations are sparsely sampled from trajectories of a phenotype contaminated with measurement error, where the phenotypic trajectory consists of a genetic component and an environmental component. By combining information across individuals, the genetic and environmental covariances are estimated via smoothing techniques. We study the genetic and environmental effects using principal component analysis, taking into account the genetic correlation to enhance the subject-level signal extraction. We show via the real data and simulations that incorporating the correlation structure improves predictions of individual phenotypic trajectories.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据