Invited review: Advances and applications of random regression models: From quantitative genetics to genomics

摘要

An important goal in animal breeding is to improvelongitudinal traits; that is, traits recorded multipletimes during an individual’s lifetime or physiologicalcycle. Longitudinal traits were first genetically evaluatedbased on accumulated phenotypic expression,phenotypic expression at specific time points, or repeatabilitymodels. Until now, the genetic evaluation oflongitudinal traits has mainly focused on using randomregression models (RRM). Random regression modelsenable fitting random genetic and environmental effectsover time, which results in higher accuracy of estimatedbreeding values compared with other statisticalapproaches. In addition, RRM provide insights abouttemporal variation of biological processes and thephysiological implications underlying the studied traits.Despite the fact that genomic information has substantiallycontributed to increase the rates of genetic progressfor a variety of economically important traits inseveral livestock species, less attention has been givento longitudinal traits in recent years. However, includinggenomic information to evaluate longitudinal traitsusing RRM is a feasible alternative to yield more accurateselection and culling decisions, because selectionof young animals may be based on the complete patternof the production curve with higher accuracy comparedwith the use of traditional parent average (i.e., withoutgenomic information). Moreover, RRM can be used toestimate SNP effects over time in genome-wide associationstudies. Thus, by analyzing marker associationsover time, regions with higher effects at specific pointsin time are more likely to be identified. Despite the advancesin applications of RRM in genetic evaluations,more research is needed to successfully combine RRMand genomic information. Future research should providea better understanding of the temporal variationof biological processes and their physiological implicationsunderlying the longitudinal traits.