Structural Variant Prediction in Extended Pedigrees Through Sparse Negative Binomial Genome Signal Recovery

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Structural variants (SVs) are genomic rearrangements with significant implications for human health.
• Accurate detection of SVs is challenging due to high false positive rates from sequencing and mapping errors.
• Previous work introduced a maximum likelihood method using low-coverage sequencing data and coverage distribution.

Purpose of the Study:

• To enhance structural variant identification accuracy by integrating parent-offspring relationships with an existing negative binomial framework.
• To reduce false positive rates in structural variant detection.
• To validate the improved method on simulated and real-world genomic data.

Main Methods:

• Developed a statistical framework leveraging parent-offspring trio data.
• Incorporated a negative binomial model to accurately represent DNA fragment distributions.
• Applied the enhanced method to simulated genomes and two parent-child trios from the 1000 Genomes Project.

Main Results:

• The integrated approach demonstrated improved accuracy in structural variant identification compared to previous methods.
• Reduced false positive rates were observed in both simulated and real genomic datasets.
• The method effectively utilized familial relationships to refine variant calls.

Conclusions:

• Integrating familial information with a negative binomial model significantly enhances structural variant detection accuracy.
• This approach offers a more robust solution for identifying genomic rearrangements, particularly in low-coverage sequencing scenarios.
• The findings have implications for improving genomic analyses in both research and clinical settings.