Page 196 - Linear Models for the Prediction of Animal Breeding Values 3rd Edition
P. 196

The equations for obtaining the solutions for SNP and polygenic effects are:
            ⎛  XX  XR X     X Z        ⎞ ⎛ ⎞ ˆ  ⎛ Xy ′  ⎞
                              ′
                     ′
               ′
                       −1
                                         b
            ⎜  ′    ′        ′         ⎟ ⎜ ⎟  ⎜   ⎟
                                         g ˆ = Zy ′
            ⎜  ZX  ZX       Z Z        ⎟ ⎜ ⎟ ⎟  ⎜  ⎟                        (11.4)
            ⎜ ⎝  WX WX      WW +  A a  ⎟ ⎜ u ˆ ⎟  ⎝ Wy ′  ⎠
               ′
                     ′
                                       ⎠ ⎝ ⎠
                                    −1
         where:
                 2  2
                 e  u
            a = s /s
            If the vector of observations, y in Eqn 11.3, are de-regressed breeding values of bulls
         (see Section 5.5.2), then each observation may be associated with differing reliabilities.
         Thus a weighted analysis may be required to account for these differences in bull reli-
         abilities. The weight (wt ) for each observation could be the reciprocal of the effective
                             i
         daughter contribution (see Section 5.5.2) or wt  = (1/rel ) – 1, where rel  is the bull’s
                                                i      dtr            dtr
         reliability from daughters with parent information excluded (VanRaden, 2008). Then the
         MME are:
            æ æ X¢RX  X¢RX     X¢R Z            öæ ö ˆ b  æ X¢ Ry  ö
                 -1
                                                          -1
                          -1
                                  -1
            ç                                   ÷ç ÷  ç       ÷ ÷
                                                          -1
            ç Z¢RX    Z¢RX    Z¢R Z             ÷ç ÷ = çZ¢ Ry ÷             (11.5)
                         -1
                                  -1
                 -1
                                                  ˆ g
            ç ç  -1         -1       -1      -1  ÷ç ÷  ç   -1  ÷
                                                øè ø
            è W¢RX      W¢RX      WR W +   A a ÷ç ÷ ˆ u  ç W¢ Ry ÷ ø
                                                      è
         where R = D and D is a diagonal matrix with diagonal element i = wt .
                                                                     i
         Example 11.1
         Given below is the real genotype for the first ten SNPs of a popular dairy bull and
         those of his sons and some other unrelated bulls genotyped using the 50K Illumin chip.
         The genotypes of animals are coded as described in Section 11.3. The observations are
         the DYDs for fat yield, and the effective daughter contribution (EDC) for each bull is
         also given. The EDC can be used as weights in the analysis. It is assumed the genetic
                                                                  2
                                      2
         variance for fat yield is 35.241 kg  and residual variance of 245 kg , and animals 13
         to 20 are assumed as the reference population and 21 to 26 as selection candidates.
         Assuming that the first three SNPs have been identified as having the most significant
         effect, the aim is to fit Eqn 11.3 with and without weights using these three SNPs:
                                           Fat
         Animal  Sire  Dam   Mean   EDC   DYD              SNP Genotype
         13       0     0      1    558    9.0  2   0   1  1   0  0   0  2   1   2
         14       0     0      1    722   13.4  1   0   0  0   0  2   0  2   1   0
         15      13     4      1    300   12.7  1   1   2  1   1  0   0  2   1   2
         16      15     2      1     73   15.4  0   0   2  1   0  1   0  2   2   1
         17      15     5      1     52    5.9  0   1   1  2   0  0   0  2   1   2
         18      14     6      1     87    7.7  1   1   0  1   0  2   0  2   2   1
         19      14     9      1     64   10.2  0   0   1  1   0  2   0  2   2   0
         20      14     9      1    103    4.8  0   1   1  0   0  1   0  2   2   0
         21       1     3      1     13    7.6  2   0   0  0   0  1   2  2   1   2
         22      14     8      1    125    8.8  0   0   0  1   1  2   0  2   0   0
         23      14    11      1     93    9.8  0   1   1  0   0  1   0  2   2   1
                                                                         Continued
          180                                                            Chapter 11
   191   192   193   194   195   196   197   198   199   200   201