Enhanced Disease Susceptible Variant Identification via Short Identity by Descent Segments

Area of Science:

• Genetics
• Genomic Medicine
• Computational Biology

Background:

• Rare diseases affect millions globally, with low diagnostic rates.
• Identity by descent (IBD) mapping aids in identifying disease variants from common ancestors.
• Current IBD detection models struggle with short IBD segments.

Purpose of the Study:

• Introduce SILO, a novel model for detecting disease-susceptible variants in both short and long IBD segments.
• Improve diagnostic yields for rare diseases by identifying previously missed variants.

Main Methods:

• SILO uses a two-stage IBD detection process.
• Stage one: Identifies long IBD segments using common variants.
• Stage two: Detects short IBD segments using rare variants and a seed-and-extend algorithm.

Main Results:

• SILO outperforms existing models in detecting variants within short IBD segments.
• SILO shows comparable performance to existing models for long IBD segments.
• Evaluated using simulated and 1000 Genomes Project data.

Conclusions:

• SILO has the potential to increase rare disease diagnostic yields.
• The model effectively identifies disease-susceptible variants in short IBD segments.
• Further improvements in short IBD segment detection precision are needed.