To study the genetic impact on a quantitative trait, molecular markers are used as predictor variables in a statistical model. This habilitation thesis elucidated challenges accompanied with such investigations. First, the usefulness of including different kinds of genetic effects, which can be additive or non-additive, was verified. Second, dependencies between markers caused by their proximity on the genome were studied in populations with family stratification. The resulting covariance matrix deserved special attention due to its multi-functionality in several fields of genomic evaluations.<eng>
