Detecting structure of haplotypes and local ancestry

Area of Science:

• Population Genetics
• Statistical Genomics
• Bioinformatics

Background:

• Understanding haplotype structure is crucial for inferring local ancestry in admixed populations.
• Existing methods face challenges in accurately modeling linkage disequilibrium at multiple scales.

Purpose of the Study:

• To introduce a novel two-layer hidden Markov model for haplotype structure detection.
• To leverage multi-scale linkage disequilibrium information for improved local ancestry inference.
• To identify genomic regions with significant deviations in local ancestry.

Main Methods:

• Developed a two-layer hidden Markov model (HMM).
• Modeled linkage disequilibrium at intra- and inter-haplotype group levels.
• Applied the model to Mexican samples from HapMap3.

Main Results:

• The proposed HMM method demonstrates superior performance compared to state-of-the-art approaches.
• Outperformance is particularly notable in regions with short ancestral track lengths.
• Identified significant local ancestry departures from the genome-wide average on chromosomes 6 and 8 in Mexican samples.

Conclusions:

• The two-layer HMM provides a powerful framework for haplotype analysis and local ancestry inference.
• The findings highlight specific chromosomal regions with unique admixture patterns in the studied population.
• A software package is available for public use.