A Linear Regression Model for the Prediction of Rice Sheath Blight Field Resistance

文献类型： 外文期刊

第一作者： Zeng, Yuxiang

作者： Zeng, Yuxiang;Dong, Junjie;Ji, Zhijuan;Yang, Changdeng;Liang, Yan

作者机构：

关键词： association mapping; field resistance; prediction model; rice; sheath blight; field crops

期刊名称：PLANT DISEASE （ 影响因子：4.614； 五年影响因子：5.33 ）

ISSN： 0191-2917

年卷期： 2021 年 105 卷 10 期

页码：

收录情况： SCI

摘要： Rice sheath blight (SB) disease is a global issue that causes great yield losses each year. To explore whether SB field resistance can be predicted, 273 rice genotypes were inoculated and evaluated for SB field resistance across nine environments from 2012 to 2019 to identify loci associated with SB resistance by association mapping. A total of 80 significant marker-trait associations were detected in nine environments, among which six loci (D130B, D230A, D304B, D309, D427A, and RM409) were repeatedly detected in at least two environments. A linear regression model for predicting SB lesion length was developed using genotypic data of these six loci and SB field resistance data of the 273 rice genotypes: y = 34.44 - 0.56x, where y is the predicted value of lesion length, and x is the total genotypic value of the six loci. A recombinant inbred line (RIL) population consisting of 219 lines that was grown in six environments (from 2013 to 2018) for evaluation of SB field resistance was used to check the prediction accuracy of the prediction model. The average absolute error between the predicted lesion length and real lesion length for the RIL population was 6.67 cm. The absolute errors between predicted and real lesion lengths were <6 cm for 51.22% of the lines and <9 cm for 71.22% of the lines. An SB visual rating prediction model was also developed, and the average absolute error between the predicted visual rating and real visual rating for the RIL population was 0.94. These results indicated that the rice SB lesion length can be predicted by the development of a linear regression model using both genotypic and phenotypic data.

分类号：

  • 相关文献
作者其他论文 更多>>

微信小程序