Partitioning of genetic effects on lifetime performance of mice.

A total of 2,457 lifetime performance records of 29 genetic groups of mice was analyzed using multiple regression of records on the proportion of gene contribution from 6 lines (designated as Lines MP, mQ, WP, wQ, CP and cQ). Genetic effects were partitioned into line additive, line maternal, direct heterosis, maternal heterosis and paternal heterosis effects. The line additive and line maternal effects were expressed as deviations from Line cQ. Seventeen of 25 line additive effects differed significantly (P<0.05) from Line cQ whereas only 4 of 25 line maternal effects deviated significantly from Line cQ. Deviations in line additive effects from cQ were negative in all lines examined whereas deviations in line maternal effects from cQ were all positive, indicating a negative relationship between line additive and line maternal effects. Direct heterosis effects were all positive and significant (P < 0.01) except in the MPxWP cross which was produced by mating Lines MP and WP of the same base population (P). Maternal heterosis effects were significant in 10 of 20 cases whereas paternal heterosis effects were significant in 13 of 20 cases. Although direct heterosis is a major component of total heterosis effects (sum of direct, maternal and paternal heterosis), the results suggest that parental heterosis may need to be considered in producing multiple way crosses. The fitting of line additive, line maternal, direct heterosis, maternal heterosis and paternal heterosis effects in the multiple regression model effectively accounted for all genetic effects in lifetime performance.