Fig. 5
Impact of data misassignment and pedigree errors on imputation accuracy. The dashed line separates the individual directly affected by the data modification (ind) and its relatives (gpar: grandparents, par: parents, prog: progeny, gprog: grandprogeny, misass prog: misassigned progeny, misass gprog: misassigned grandprogeny). The y-axis measures the individual-wise dosage correlation between the imputed genotypes based on complete correct data and either missing or misassigned data for the individual itself and its relatives. In panel a we provide the case where the sequence data of the target individual was masked as in Test 1; in panel b where the sequence data of another individual was misassigned to the target one; in panel c where the marker array data was misassigned; and in panel d where we assigned the progeny from one of the individuals sequenced at high coverage to the target individual