Probabilistic method for detecting copy number variation in a fetal genome using maternal plasma sequencing

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Maternal plasma contains fetal DNA, enabling non-invasive prenatal testing.
• Current methods detect chromosomal abnormalities and larger copy number variants (CNVs).
• Smaller fetal CNVs remain challenging to detect non-invasively.

Purpose of the Study:

• To develop a novel probabilistic method for analyzing de novo CNVs in the fetal genome from maternal plasma.
• To improve the detection accuracy and precision of fetal CNVs.

Main Methods:

• Utilized a Hidden Markov Model integrating allelic ratio imbalance, parental genotypes, and depth of coverage.
• Developed a method for non-invasive analysis of de novo CNVs in fetal genomes.
• Simulated data with known CNVs introduced into maternal plasma samples.

Main Results:

• Achieved 90% sensitivity for CNVs >400 kb with minimal false positives.
• Demonstrated 40% sensitivity for smaller CNVs (50-400 kb).
• The method effectively differentiates various CNV types and enhances precision.

Conclusions:

• The developed probabilistic method offers improved non-invasive detection of fetal CNVs.
• This advancement has potential implications for prenatal genetic screening.
• The computational model and simulation methods are publicly available.