In addition to potentially discovering sample handling issues, visualizing sample relatedness as shown in Figure 4 also reveals any cryptic relatedness that may be present in the study sample. Figure 4 shows that many individuals who indicated that they were unrelated (black points) or distantly related (blue points) line up along the diagonal in this plot. These individuals represent second, third, fourth, and fifth degree relatives. If treated as independent samples in the downstream analyses, having many related samples in the dataset would result in increased type I and type II errors, thus analytical methods such as mixed model regression [22] must be used in place of simple linear or logistic regression. Figure 5 shows another way to visualize the degree of relatedness by plotting a histogram of the distribution of kinship coefficients over 0.05 between all pairs of individuals in the dataset.
