Including the genetic values of the possible parents when analyzing pedigrees that contain MULT groups.

Abstract

In a studbook, MULT is used in a parent ID field when the actual parent is unknown but the parent is known to be one of a set of possible parents. Probabilities of being the actual parent are assigned to each possible parent in the MULT group, and that information is used in the calculation of mean kinships (MKs). Parental probabilities are typically assigned based on the species biology and/or what was known about how the animals were being managed at the time of conception. If there is no additional information, the default is to assign each possible parent the same probability. What has not been considered to date is the impact of different MKs among the group of possible parents. Methods are developed which allow a combination of parental probabilities and MKs into parental weights. These weights replace parental probabilities in the analysis. One important conclusion is that even when the MKs of possible parents are quite different, the difference between the parental weights and probabilities is typically less than 30%. This highlights the importance of correct estimation of parental probabilities, whenever possible, instead of reliance on a default.