Comparison of models for genetic evaluation of number of inseminations to conception in Danish Holstein cows

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review

Standard

Comparison of models for genetic evaluation of number of inseminations to conception in Danish Holstein cows. / Guo, Gang; Hou, Yali; Zhang, Yuan; Su, Guosheng.

In: Animal Science Journal, Vol. 88, No. 4, 2017, p. 567-574.

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review

Harvard

APA

CBE

MLA

Vancouver

Author

Guo, Gang ; Hou, Yali ; Zhang, Yuan ; Su, Guosheng. / Comparison of models for genetic evaluation of number of inseminations to conception in Danish Holstein cows. In: Animal Science Journal. 2017 ; Vol. 88, No. 4. pp. 567-574.

Bibtex

@article{06199f5e47b64932a3bb55ef21d89849,
title = "Comparison of models for genetic evaluation of number of inseminations to conception in Danish Holstein cows",
abstract = "Number of inseminations to conception (NINS), an important fertility trait, requires appropriate approaches for genetic evaluation due to its non-normal distribution and censoring records. In this study, we analyzed NINS in 474837 Danish Holstein cows at their first lactation by using seven models which deal with the categorical phenotypes and censored records in different manners, further assessed these models with regard to stability, lack of bias and accuracy of prediction. The estimated heritability from four models based on original NINS specified as a linear Gaussian model, categorical threshold model, threshold linear model and survival model were similar (0.031-0.037). While for the other three models based on the binary response derived from NINS, referred as threshold model (TM), logistic and probit models (LOGM and PROM), the heritability were estimated as 0.027, 0.063 and 0.027, respectively. The model comparison concluded that different models could lead to slightly different sire rankings in terms of breeding values; a more complicated model led to less stability of prediction; the models based on the binary response derived from NINS (TM, LOGM and PROM) had slightly better performances in terms of unbiased and accurate prediction of breeding values.",
keywords = "Female fertility, Genetic evaluation, Holstein cow, Model comparison, Number of insemination to conception",
author = "Gang Guo and Yali Hou and Yuan Zhang and Guosheng Su",
year = "2017",
doi = "10.1111/asj.12406",
language = "English",
volume = "88",
pages = "567--574",
journal = "Animal Science Journal",
issn = "1344-3941",
publisher = "Wiley-Blackwell Publishing Asia",
number = "4",

}

RIS

TY - JOUR

T1 - Comparison of models for genetic evaluation of number of inseminations to conception in Danish Holstein cows

AU - Guo, Gang

AU - Hou, Yali

AU - Zhang, Yuan

AU - Su, Guosheng

PY - 2017

Y1 - 2017

N2 - Number of inseminations to conception (NINS), an important fertility trait, requires appropriate approaches for genetic evaluation due to its non-normal distribution and censoring records. In this study, we analyzed NINS in 474837 Danish Holstein cows at their first lactation by using seven models which deal with the categorical phenotypes and censored records in different manners, further assessed these models with regard to stability, lack of bias and accuracy of prediction. The estimated heritability from four models based on original NINS specified as a linear Gaussian model, categorical threshold model, threshold linear model and survival model were similar (0.031-0.037). While for the other three models based on the binary response derived from NINS, referred as threshold model (TM), logistic and probit models (LOGM and PROM), the heritability were estimated as 0.027, 0.063 and 0.027, respectively. The model comparison concluded that different models could lead to slightly different sire rankings in terms of breeding values; a more complicated model led to less stability of prediction; the models based on the binary response derived from NINS (TM, LOGM and PROM) had slightly better performances in terms of unbiased and accurate prediction of breeding values.

AB - Number of inseminations to conception (NINS), an important fertility trait, requires appropriate approaches for genetic evaluation due to its non-normal distribution and censoring records. In this study, we analyzed NINS in 474837 Danish Holstein cows at their first lactation by using seven models which deal with the categorical phenotypes and censored records in different manners, further assessed these models with regard to stability, lack of bias and accuracy of prediction. The estimated heritability from four models based on original NINS specified as a linear Gaussian model, categorical threshold model, threshold linear model and survival model were similar (0.031-0.037). While for the other three models based on the binary response derived from NINS, referred as threshold model (TM), logistic and probit models (LOGM and PROM), the heritability were estimated as 0.027, 0.063 and 0.027, respectively. The model comparison concluded that different models could lead to slightly different sire rankings in terms of breeding values; a more complicated model led to less stability of prediction; the models based on the binary response derived from NINS (TM, LOGM and PROM) had slightly better performances in terms of unbiased and accurate prediction of breeding values.

KW - Female fertility

KW - Genetic evaluation

KW - Holstein cow

KW - Model comparison

KW - Number of insemination to conception

UR - http://www.scopus.com/inward/record.url?scp=84982883888&partnerID=8YFLogxK

U2 - 10.1111/asj.12406

DO - 10.1111/asj.12406

M3 - Journal article

C2 - 27506328

AN - SCOPUS:84982883888

VL - 88

SP - 567

EP - 574

JO - Animal Science Journal

JF - Animal Science Journal

SN - 1344-3941

IS - 4

ER -