Genotype imputation reference panel selection using maximal phylogenetic diversity.

Area of Science:

• Genomics
• Bioinformatics
• Population Genetics

Background:

• Next-generation sequencing (NGS) costs are decreasing, enabling genome-wide variant assessment for complex diseases.
• Sequencing large sample sizes remains prohibitively expensive.
• Imputation using a sequenced subset as an internal reference panel is a cost-effective strategy for samples with existing genotype data.

Purpose of the Study:

• To develop a cost-effective strategy for genetic imputation in large studies with existing genotype data.
• To introduce and evaluate a novel method for selecting optimal internal reference panels.
• To improve imputation accuracy, particularly for rare variants.

Main Methods:

• Introduced the concept of phylogenetic diversity from mathematical phylogenetics.
• Proposed the "most diverse reference panel" selection strategy based on maximal phylogenetic diversity.
• Validated the method using simulations and data from the 1000 Genomes Project.

Main Results:

• The "most diverse reference panel" significantly improved imputation accuracy compared to random selection.
• This improvement was particularly pronounced for the imputation of rare variants.
• Enhanced accuracy was observed across various marker densities, reference panel sizes, and imputed segment lengths.

Conclusions:

• The "most diverse reference panel" strategy offers a substantial improvement in imputation accuracy.
• This approach is effective for identifying population-specific variants and mitigating ancestry mismatch.
• The proposed method provides a novel, cost-effective strategy for planning sequencing studies.