Skip to main content
. 2019 Oct 10;105(5):921–932. doi: 10.1016/j.ajhg.2019.09.016

Figure 4.

Figure 4

Aneuploidy Rates and Maternal Age Are Correlated with UPD

(A) The correlation between per-chromosome UPD rate in true positives from the 23andMe database and per chromosome aneuploidy rate in published pre-implantation genetic screening data;4 chromosomes are colored by centromeric type: metacentric chromosomes are shown in red, submetacentric chromosomes in green, and acrocentric chromosomes in blue. These two rates are significantly correlated (Pearson’s correlation = 0.49; p value = 0.02) and this correlation remains significant after correction for chromosome length and centromeric type (Pearson’s correlation = 0.73; p value = 0.006), suggesting that meiotic nondisjunction occurs more frequently on some chromosomes (such as 15, 16, 21, and 22) than others, resulting in more instances of both UPD and aneuploidy on these chromosomes. We also note that the acrocentric chromosomes have among the highest per-chromosome rates of both UPD and aneuploidy.

(B) The age distribution of mothers of UPD true negatives (blue) and that of mothers who are parents of origin of UPD true positives (matUPD cases, yellow) in the 23andMe dataset. We find that mothers of UPD case subjects are significantly older than mothers of UPD true negatives (Wilcoxon p value = 0.00948) and that this associations holds when restricted to cases of matUPD, where mothers are the parents of origin of the UPD case subjects (Wilcoxon p value = 0.00317).