| 摘要: |
| 将广义线性混合模型（GLMM）技术引入家畜阈性状的遗传分析,对不同参数组合下阈性状的遗传力及准确度进行了估计,模拟研究的性状为单阈值和3阈值离散性状,试验设计为全同胞-半同胞混合家系.结果表明,GLMM方法能有效地估计阈性状的遗传力,其准确性有较大的优势.不同参数组合下,单阈值与3阈值分类性状的遗传力估计均方误（MSE）均在0.018 4以下;遗传力参数的估计偏差（BIAS）均小于0.429 0.同时,性状的遗传力和性状的表型发生率对遗传力估计效果都有直接影响,随着性状遗传力真值的提高,GLMM方法出现高估现象;随着性状发生率的提高,GLMM方法估计的遗传力结果有上升的趋势.说明利用GLMM方法估计家畜阈性状的遗传力是可行的. |
| 关键词: 广义线性混合模型（GLMM） 阈性状 遗传力 遗传分析 |
| DOI：10.11841/j.issn.1007-4333.2005.06.154 |
| 投稿时间：2005-09-01 |
| 基金项目:国家重点基础研究发展规划资助项目（G2000016103）;安徽省自然科学基金资助项目（050410204）;安徽省教育厅资助项目（2002jq126,2004kj151） |
|
| Estimation of heritability for threshold traits in livestock using GLMM method |
|
|
|
| Abstract: |
| The method of generalized linear mixed models(GLMM) was described.It allowed genetic analyses and heritability efficiency estimate for threshold traits in livestock under different parameters.The results showed that GLMM is advantageous in heritability estimation and accurate for discrete traits.The estimated maximal mean standard error(MSE) and heritability bias(BIAS) are respectively 0.018?4 and 0.429?0 for all different parameters.The increase of the true heritability and incidence of categorie induced the improvement of heritability accuracy,showing that the accuracy of estimation directly depended on these two parameters. |
| Key words: generalized linear mixed model(GLMM),threshold traits,heritability,genetical analysis |
