LowKi

LowKi estimates relatedness coefficients from low-depth whole-genome sequencing (WGS) data to enable inference of kinship and fraternity coefficients when individual genotypes cannot be confidently called.

Key Features:

• Method-of-Moment Estimators: Implements novel method-of-moment estimators that compute relatedness directly from genotype likelihoods rather than relying on called genotypes.
• Genotype Likelihoods: Uses genotype likelihoods to account for genotype uncertainty inherent to low-depth sequencing.
• Targets Kinship Metrics: Provides estimates specifically for kinship and fraternity coefficients relevant to pedigree and relatedness analyses.
• Robustness and Efficiency: Demonstrates robustness compared to alternative approaches and is optimized for computational efficiency and minimal external dependencies.

Scientific Applications:

• Population Genetics: Enables relatedness estimation in population-genetic studies using low-depth WGS data.
• Family-Based Association Studies: Supports accurate kinship estimation for family-based association analyses where genotype calls are uncertain.
• Evolutionary Biology: Facilitates inference of familial relationships and relatedness patterns in evolutionary research.
• Large-Scale Low-Depth Datasets: Allows analysis of large cohorts sequenced at low depth by providing reliable relatedness estimates without confident genotype calls.

Methodology:

Computations use novel method-of-moment estimators applied to genotype likelihoods; low-depth data were simulated using a complex pedigree from the Cilento isolates in South Italy; validation was performed on a sample of 150 French individuals with down-sampled sequencing data.