Page 84 - Linear Models for the Prediction of Animal Breeding Values 3rd Edition
P. 84

Piglet     Sire     Dam       Sex         Weaning weight (kg)
                 6          1        2        Male               90
                 7          1        2        Female             70
                 8          1        2        Female             65
                 9          3        4        Female             98
                10          3        4        Male              106
                11          3        4        Female             60
                12          3        4        Female             80
                13          1        5        Male              100
                14          1        5        Female             85
                15          1        5        Male               68

         The objective is to predict breeding values for all animals and common environmental
                                                                  2
                                    2
                                                      2
                                            2
         effects for full-sibs. Given that s  = 20, s  = 15 and s  = 65, then s  = 100, a  = 3.25
                                    a       c         e           y        1
         and a  = 4.333.
              2
            The model for the analysis is that presented in Eqn 4.5 and, as mentioned earlier,
         the MME for the BLUP of a and c and BLUE of estimable functions of b are as given
         in Eqn 4.2, using a  and a  defined above.
                         1      2
         SETTING UP THE DESIGN MATRICES
         The transpose of the matrix X, which relates records to sex effects in this example is:
                ⎡1 000 1 0 0 1 01           ⎤
            X ′ =  ⎢                        ⎥
                ⎣ 0 1 1 1 0 1 10 10         ⎦
         and Z =  I, excluding parents. The transpose of matrix  W that relates records to
         full-sibs is:

                  ⎡111 000000             ⎤
                  ⎢                       ⎥
            W ′ = 000 111 000             ⎥
                  ⎢
                  ⎣ ⎢000000 111           ⎥ ⎦

            The MME can be set up as discussed in Example 4.1. The solutions to the MME
         by direct inversion of the coefficient matrix are:

                             Effects                  Solutions

                             Sex
                               Male                    91.493
                               Female                  75.764
                             Animals
                               1                       −1.441
                               2                       −1.175
                               3                        1.441
                               4                        1.441
                               5                       −0.266
                               6                       −1.098
                                                     Continued


          68                                                              Chapter 4
   79   80   81   82   83   84   85   86   87   88   89