BalLeRMix

BalLeRMix applies mixture-model-based composite likelihood ratio tests (Bstatistics) to detect long-term balancing selection in genomic datasets.

Key Features:

• Bstatistics suite: A suite of five composite likelihood ratio test statistics derived from mixture models for detecting balancing selection.
• Window-agnostic approach: Bstatistics operate independently of genomic window size, reducing the trade-off between noise and statistical power across varying window dimensions.
• Robustness to mutation and recombination: Demonstrated resilience to high mutation rates and uneven recombination landscapes that can confound population-genomic analyses.
• Versatile input compatibility: Bstatistics can process a wide range of input data forms for application to diverse genomic datasets.
• Multi-allelic selection extension: Methodology extended to account for multi-allelic balancing selection scenarios.

Scientific Applications:

• Human population-genomic analysis: Applied to human datasets to identify candidate genes including STPG2, CCDC169-SOHLH2, KLRD1, and SCN9A.
• Bonobo population-genomic analysis: Applied to bonobo datasets to confirm candidates in the MHC-DQ region and reveal novel candidates related to immune response and sensory functions.

Methodology:

The method employs mixture models to compute a suite of five composite likelihood ratio test statistics (Bstatistics) and includes extensions for multi-allelic balancing selection.