Inference of relationships in population data using identity-by-descent and identity-by-state

Area of Science:

• Population Genomics
• Human Genetics
• Bioinformatics

Background:

• Accurate sample annotation is crucial for population-based genetic studies.
• Existing relatedness assessment methods rely on identity-by-descent (IBD) and identity-by-state (IBS) allele-sharing proportions.
• A gap exists in practical, scalable methods for analyzing large datasets without prior relatedness information.

Purpose of the Study:

• To develop a novel approach for estimating identity-by-descent (IBD0, IBD1, and IBD2) from observed identity-by-state (IBS) within genomic windows.
• To provide an intuitive and practical graphical method for analyzing large datasets with thousands of samples.
• To accurately identify related individuals, detect inbreeding, and pinpoint population outliers.

Main Methods:

• Developed a novel method estimating IBD0, 1, and 2 based on observed IBS within windows.
• Combined genome-wide IBS information with the novel IBD estimation for graphical analysis.
• Applied the method to a Human Variation Panel and benchmarked against PLINK's hidden Markov model.

Main Results:

• Identified identical, parent-child, and full-sibling relationships, reconstructing pedigrees within a nominally unrelated panel.
• Detected unexpected identity-by-descent (IBD2) levels and homozygosity regions, indicating inbreeding in some pairs.
• Successfully distinguished related individuals from those with atypical heterozygosity and identified population outliers.
• Identified distant relatedness within populations using megabase-scale regions lacking IBS0, outperforming PLINK in calling distantly related individuals.

Conclusions:

• The novel IBD estimation method offers an intuitive and practical approach for analyzing large population datasets.
• This method improves the accuracy of relatedness assessment, inbreeding detection, and outlier identification in genomic studies.
• The approach has broad applications in genome-wide association, linkage, heterozygosity, and other population genomics studies relying on SNP genotype data.