Whole exome sequencing of mother could predict risk of creating embryos with aneuploidy

Variants in three genes have been linked to an increased risk of a woman creating aneuploid embryos while undergoing IVF and used to create a risk score using machine learning.

Embryonic aneuploidy describes an embryo with an abnormal number of chromosomes. This abnormality can lead to increased risk of miscarriage. The risk of producing aneuploid embryos increases with older maternal age as the eggs are more likely to be aneuploid. However, little is known about the cause of egg aneuploidy so researchers designed a study to determine if studying the genome of the mother could help indicate the risk of embryonic aneuploidy. Knowing this could help couples determine their chances of successful IVF with their own eggs, authors of the paper published in Human Genetics argued.

Dr Jinchuan Xing, research lead author and associate professor at the Rutgers School of Arts and Sciences, New Brunswick, New Jersey, said: 'The goal of our project was to understand the genetic cause of female infertility and develop a method to improve clinical prognosis of patients' aneuploidy risk. Based on our work, we showed that the risk of embryonic aneuploidy in female IVF patients can be predicted with high accuracy with the patients' genomic data. We also have identified several potential aneuploidy risk genes.'

First, researchers analysed the rate of embryo aneuploidy in women undergoing IVF in order to determine their individual risk. They selected 281 women of European ancestry from two genomic datasets. They then  them into two groups: a low rate group if they had less than 30 percent rate of aneuploidy in their embryos and a high rate group if it was over 50 percent. Then they analysed the women's genomic data to determine whether there were any genetic variants linked to a higher risk of creating embryos with aneuploidy.

This flagged up 23 genes associated with meiosis, which researchers used to develop a risk score for embryonic aneuploidy. They then developed machine learning models to test whether or not presence of these genetic variants could be used to determine unseen women's risk of creating embryos with aneuploidy using a different cohort to test their model. Variants on three genes MCM5, FGGY, and DDX60L were found to contribute the most to the model's predictive power, and could be used as future targets in diagnostic or therapeutic approaches.

Researchers said the risk score could be more useful than just looking at maternal age, as aneuploidy rates can vary considerably between individuals.

'I like to think of the coming era of genetic medicine when a woman can enter a doctor's office or, in this case, perhaps, a fertility clinic with her genomic information, and have a better sense of how to approach treatment. Our work will enable such a future.' Dr Xing said.