Journal article
MultiBLUP: Improved SNP-based prediction for complex traits
D Speed, DJ Balding
Genome Research | Published : 2014
Abstract
BLUP (best linear unbiased prediction) is widely used to predict complex traits in plant and animal breeding, and increasingly in human genetics. The BLUP mathematical model, which consists of a single random effect term, was adequate when kinships were measured from pedigrees. However, when genome-wide SNPs are used to measure kinships, the BLUP model implicitly assumes that all SNPs have the same effect-size distribution, which is a severe and unnecessary limitation. We propose MultiBLUP, which extends the BLUP model to include multiple random effects, allowing greatly improved prediction when the random effects correspond to classes of SNPs with distinct effect-size variances. The SNP cla..
View full abstractGrants
Awarded by National Institute for Health Research
Funding Acknowledgements
We thank David van Heel of Queen Mary University of London for providing the celiac disease data, Sang Lee of the Queensland Institute of Medical Research for helpful advice regarding average information REML, and three anonymous reviewers for their constructive suggestions. Analyses were performed with the use of the UCL Legion High Performance Computing Facility (Legion@UCL) and with the help of the associated support services. Access to Wellcome Trust Case Control Consortium data was authorized as work related to the project "Genome wide association study of susceptibility and clinical phenotypes in epilepsy"; and access to data from the National Institute of Diabetes and Digestive and Kidney Disease was granted under Project 5938, "Using genome-wide SNP data to predict disease behavior for Crohn's disease." This work is funded by the UK Medical Research Council under grant G0901388, with support from the National Institute for Health Research, University College London Hospitals Biomedical Research Centre.