估计遗传配合力的线性规划法

LINEAR PROGRAMMING METHOD OF ESTIMATING GENETIC COMBINING ABILITY

本文提出配合力模型：Ｙ＿（ｉｊｋ１）＝μ＋ｆ＿ｉ＋ｍ＿ｉ＋ｓ＿（ｉｊ）＋ｒ＿ｋ＋ｆr＿（ｉｋ）＋ｍｒ＿（ｊｋ）＋ｓｒ＿（ｉｊｋ）＋ｅ＿（ｉｊｋｌ），这里ｆ＿ｉ、ｍ＿ｊ为父母的一般配合力，ｓ＿（ｉｊ）为特殊配合力，ｒ＿ｋ为环境效应，ｆｒ＿（ｉｋ）、ｍｒ＿（ｊｋ）为遗传环境互作效应，ｓｒ＿（ｉｊｋ）为特殊配合力与环境的互作效应，ｅ＿（ｉｊｋｌ）为误差。依据最小二乘原理，可由试验观察数据建立约束方程，用配合力和互作效应估计误差总和来建立目标函数，在令估计误差总和为最小的条件下，取得配合力和效应值的最佳估计。这样由于观察误差与效应值误差都取最小值，可使估计值更为可靠，且可以避免令效应值总和为零造成的问题。适合于在样本含量不等时，对配合力进行估计；在选配和杂交中可用于决定选配组合和合适的饲养管理条件。

It is to advance in this paper the model of combining ability:y_ijk1=μ+f_i+m_j+s_ij+r_k+fr_ik+mr_jt+sr_ijk+e_ijk1 where f_i and m_j mean parents' general combining ability,s_ij the special combiningability,r_k the environmental effect,fr_ik and mr_jk genotype and environment interaction,sr_ijk the interation between special combining ability and environment and e_ijk1 theerror.According to the least square principle,it's possible to establish the restrictedequation with the observcd data,and to set the objective function with estimated errorsummation by combining ability and interaction.To make the estimated error summationthe smallest,we can gain the best estimation of combining ability and effect valtie,thus the estimated value is more reliable due to the smallest value of the observederror and effect value error.Moreover,it's possible to avoid the problem of making thesummation of effect value zero.It's suitable for estimating the colnbining ability whenthe subsamples numbers are unequal,and it can be used to determine the cross combi-oation and feeding management.