Detecting inherited and novel structural variants in low-coverage parent-child sequencing data.

Area of Science:

• Genomics and Bioinformatics
• Human Genetics
• Computational Biology

Background:

• Structural variants (SVs) are significant genomic variations associated with diseases and cancer susceptibility.
• Current SV detection methods rely on mapping genome fragments to a reference, often producing spurious predictions due to sequencing and mapping errors, especially at low coverage.
• Previous SV detection methods using parent-child trios assumed all child variants were inherited, potentially missing rare, novel variants.

Purpose of the Study:

• To generalize existing SV detection methods to accommodate novel variants present only in a child's genome.
• To improve the accuracy and scope of structural variant detection by considering both inherited and novel variants.
• To enhance variant prediction in parent-child genomic comparisons.

Main Methods:

• Developed a generalized approach for SV detection in parent-child pairs, specifically allowing for novel variants in the child.
• Employed a constrained optimization framework to differentiate between inherited and novel variants.
• Tested the method on simulated data, parent-child trios from the 1000 Genomes Project, and Platinum Genomes data.

Main Results:

• The generalized approach demonstrated improved variant prediction capabilities compared to previous methods.
• The method successfully identified both inherited and novel structural variants in parent-child genomic data.
• Validation on real-world datasets confirmed the enhanced power of the new approach.

Conclusions:

• The developed method effectively improves structural variant detection by incorporating the possibility of novel variants in offspring.
• This approach offers a more comprehensive analysis of genomic variation in family trios.
• The findings contribute to more accurate identification of genetic variants underlying diseases and cancer.