A method for detecting IBD regions simultaneously in multiple individuals--with applications to disease genetics

Area of Science:

• Genetics and Genomics
• Computational Biology
• Population Genetics

Background:

• Individuals in a population share genomic regions identical by descent (IBD) due to common ancestry.
• Detecting IBD regions is crucial for understanding human evolution and identifying disease-related genetic variants.
• Existing methods for IBD detection often require pedigree information or are limited to pairwise comparisons.

Purpose of the Study:

• To develop a novel, non-pedigree based method for detecting identical by descent (IBD) regions.
• To enable simultaneous inference of IBD sharing among multiple individuals.
• To improve the accuracy and power of IBD detection using probabilistic models for unphased SNP data.

Main Methods:

• A new Markov Chain Monte Carlo (MCMC) method was developed for IBD region detection.
• The method utilizes a probabilistic model applicable to unphased single nucleotide polymorphism (SNP) data.
• It accounts for factors including inbreeding, allele frequencies, genotyping errors, and genomic distances.

Main Results:

• Simulations demonstrated that simultaneous modeling of multiple individuals enhances the power and accuracy of IBD detection compared to existing non-pedigree methods.
• The method successfully mapped a known disease-causing mutation to a 2.2-Mb region using SNP data from only five affected individuals.
• This fine-mapping capability surpasses limitations of classical linkage and association mapping techniques.

Conclusions:

• The developed MCMC method provides a powerful and accurate approach for detecting IBD regions without pedigree information.
• Simultaneous inference of IBD sharing across multiple individuals significantly improves detection capabilities.
• This method holds promise for genetic studies, particularly in mapping disease-causing variants in populations with limited family data.