Combined Model-Based Prediction for Non-Invasive Prenatal Screening.

Area of Science:

• Genetics
• Prenatal Diagnostics
• Bioinformatics

Background:

• Maternal age is a known risk factor for chromosomal abnormalities in offspring.
• Non-invasive prenatal testing (NIPT) is a widely used screening method, but its accuracy can be affected by various factors.
• Improving the precision of NIPT is crucial for accurate prenatal risk assessment.

Purpose of the Study:

• To enhance the accuracy and robustness of chromosomal abnormality prediction in NIPT.
• To evaluate the effectiveness of combining multiple analytical methods for NIPT.
• To develop a more reliable approach for prenatal screening.

Main Methods:

• Utilized whole-genome sequencing data from 1698 reference and 109 test samples.
• Applied three distinct analytical methods: standard Z-score, normalized chromosome value, and within-sample reference bin.
• Developed a logistic regression model integrating the results from the three methods.

Main Results:

• The combined logistic regression model demonstrated superior accuracy in predicting chromosomal abnormalities compared to any single method.
• Individual methods showed varying performance, but their integration yielded a significant improvement.
• The proposed combined approach proved effective in a cohort of 109 test samples.

Conclusions:

• Combining standard Z-score, normalized chromosome value, and within-sample reference bin methods enhances NIPT accuracy.
• The developed logistic regression model offers a more reliable prediction of chromosomal abnormalities.
• This integrated approach represents a promising advancement for increasing the overall reliability of non-invasive prenatal testing.