Bonsai: An efficient method for inferring large human pedigrees from genotype data.

Area of Science:

• Genetics
• Computational Biology
• Bioinformatics

Background:

• Pedigree inference from genotype data is complex, especially with sparse sampling and distant relationships.
• Current methods struggle with large or distantly related pedigrees.

Purpose of the Study:

• To develop an accurate and efficient method for inferring large pedigrees from genotype data.
• To overcome limitations of existing pedigree inference techniques for challenging datasets.

Main Methods:

• Infers small, close-kin pedigrees first, then assembles them into larger structures.
• Introduces novel formulas for relatedness degree, background identity-by-descent (IBD) detection, and ancestral branch identification.
• Incorporates age information into likelihood calculations and uses a heuristic branch-and-bound approach for efficient exploration.

Main Results:

• Successfully constructs large pedigrees that are difficult for current methods.
• Improves accuracy and efficiency in pedigree inference.
• Enables identification of distant relatives and their connecting ancestral lines.

Conclusions:

• The developed method significantly advances the capability of pedigree inference from sparse genotype data.
• It provides a robust framework for reconstructing complex family structures.
• This approach is crucial for genetic studies involving distantly related individuals.