Page 178 - Linear Models for the Prediction of Animal Breeding Values 3rd Edition
P. 178

where Q = (I − P′) and P is a matrix that relates the effect of the MQTL allele of
         an individual to the paternal and maternal MQTL alleles of its parent. Each row
         of P contains only two non-zero elements if the parent is known, otherwise only
         zeros if the parent is unknown. For instance, for individual i with sire (s) known,
                                                                    p
                             p
             p
                                                             p
         row i  will have (1 − r ) in the column corresponding to i , and r  in the column
             o               o                               s      o
                                                                                m
                                                               m
                               m
         corresponding to column i . Similarly, if dam (d) is known, row i  will contain (1 − r )
                               s                               o                o
                                                                            m
                                             m
                                      p
         in the column corresponding to i  and r  in the column corresponding to i . The
                                      d      o                              d
         row of P for allele i is equal to s  in Eqn 10.12. The matrix P for the pedigree in
                                      i
         Example 10.1 is:
                1p     1m     2p     2m     3p     3m     4p     4m     5p     5m
         1p     0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0
         1m     0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0
         2p     0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0
         2m     0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0
         3p     0.9    0.1    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0
         3m     0.0    0.0    0.1    0.9    0.0    0.0    0.0    0.0    0.0    0.0
         4p     0.1    0.9    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0
         4m     0.0    0.0    0.0    0.0    0.9    0.1    0.0    0.0    0.0    0.0
         5p     0.0    0.0    0.0    0.0    0.0    0.0    0.9    0.1    0.0    0.0
         5m     0.0    0.0    0.0    0.0    0.9    0.1    0.0    0.0    0.0    0.0
            The matrix H is a diagonal matrix for the covariance of residual effects after
         adjusting the effect of the MQTL allele of an individual for the effects of the parent’s
                                                                      p
         paternal and maternal MQTL alleles. For example, the residual effect (e ) for a pater-
                                                                      o
         nal MQTL allele of an individual with sire s known is:
                                p m
                 p
             p
                            p
                          p
            e  = v  − (1 − r )v  + r v
             o   o        o  s  o s
                           p
         and the variance of e  is:
                           o
                         p
                                                                     p
                                                                  p
                                                                               m
                                                                            p
                                          p
                                  p 2
                                                        m
                                               p 2
                 p
            var(e ) = var(v ) − (1 − r )  · var(v ) − (r )  · var(v ) − 2(1 − r )r  · cov(v , (v ))
                o       o        o        s    o        s         o  o      s  s
                                             p
                                               m
                                    2
                               m
                                                                        p
                 p
                                                       p
                        p
                                                                 m
                                                             p
                                                                               2
         Since var(v ) = var(v ) = var(v ) = s  and cov(v , v ) = var(v ) · P(Q ≡ Q ) = var(v ) · F =s F,
                 o      s      s    v        s  s      s    s    s      s  s   v s
         the above equation can be written as:
                  p
                                   p 2
                                          2
                                                   p
                                                p
                                 2
                          p
                        2
             var(e ) = 2s (r ) − 2s (r )  − 2s (1 − r )r F
                  o     v  o     v  o     v     o  o  s
                                            p
                                    p
                                               p
                           2
                                 p
                      = 2s ((1 − r )r  − (1 − r )r F )
                          v      o  o       o  o s
                           2
                                 p
                                   p
                      = 2s (1 − r )r (1 − F )
                          v      o  o     s
                  p
                     2
                         p
                                  p
                                     p
             var(e )/s  = h  = 2(1 − r )r (1 − F )                         (10.14)
                  o  v   o        o  o     s
                                      p
                       p
                    p
         where (1 − r )r   = (1 − r)r for r  = r or (1 − r), F  is the inbreeding coefficient at
                    o  o              o              s
         the MQTL of the sire and h  is the diagonal element of H for the paternal MQTL
                                  p
                                  o
         of individual o. Therefore, if the sire is not inbred, h  = 2(1 − r)r with marker infor-
                                                      p
                                                      o
         mation or h  = 0.5 with no marker information and h  = 1 if the sire is unknown.
                   p
                                                         p
                   o                                    o
         Similarly, for the maternal MQTL of o:
                     2
                                     m
                                  m
                         m
            var(e )/s  = h  = 2(1 − r )r (1 − F )                          (10.15)
                 m
                 o   v   o        o  o     d
          162                                                            Chapter 10
   173   174   175   176   177   178   179   180   181   182   183