## Abstract

Heritability is of major importance in plant studies to help achieve better yield and other agronomic traits of

interest. In candidate gene studies regression models are used to test for associations between phenotype and

candidate single nucleotide polymorphisms (SNPs). SNP imputation guarantees that marker information

is complete and so both the coefficient of determination, R 2 , and broad-sense heritability are equivalent.

However, when the normality assumption is violated, the classical R 2 may be seriously affected. Recently

two R 2 alternatives with good properties were proposed for the linear mixed model: a marginal R 2 m for the

variance explained by the fixed factors and a conditional R 2 c for the variance explained by both the fixed and

random factors. In this work we step forward a robust version of R 2 c and assess the adequacy of both classical

and robust counterparts in the estimation of true broad-sense heritability via simulation, where a particular

contamination scenario is considered. An example of application with a real maize data set is also presented.

interest. In candidate gene studies regression models are used to test for associations between phenotype and

candidate single nucleotide polymorphisms (SNPs). SNP imputation guarantees that marker information

is complete and so both the coefficient of determination, R 2 , and broad-sense heritability are equivalent.

However, when the normality assumption is violated, the classical R 2 may be seriously affected. Recently

two R 2 alternatives with good properties were proposed for the linear mixed model: a marginal R 2 m for the

variance explained by the fixed factors and a conditional R 2 c for the variance explained by both the fixed and

random factors. In this work we step forward a robust version of R 2 c and assess the adequacy of both classical

and robust counterparts in the estimation of true broad-sense heritability via simulation, where a particular

contamination scenario is considered. An example of application with a real maize data set is also presented.

Original language | English |
---|---|

Pages | 1736-1741 |

Number of pages | 6 |

Publication status | Published - Jul 2015 |

Event | 60th ISI World Statistics Congress - Riocentro, Rio de Janeiro, Brazil Duration: 26 Jul 2015 → 31 Jul 2015 http://www.isi2015.org/ |

### Conference

Conference | 60th ISI World Statistics Congress |
---|---|

Country | Brazil |

City | Rio de Janeiro |

Period | 26/07/15 → 31/07/15 |

Internet address |

## Keywords

- Robust linear mixed model
- Coefficient of determination
- Single nucleotide polymorphism
- Heritability estumation