Population genetic analysis of bi-allelic structural variants from low-coverage sequence data with an...

Area of Science:

• Population genetics
• Genomics
• Bioinformatics

Background:

• Population genetics studies typically rely on pre-genotyped variable sites.
• Accurate genotyping is challenging with low-coverage sequencing data, especially for structural variants.
• Existing statistical methods are limited to nucleotide variation and assume equal allele sampling, which is often false for structural variants.

Purpose of the Study:

• To develop a statistical method for analyzing population genetics of structural variants from low-coverage sequencing data.
• To address the limitations of current methods that require accurate genotyping and assume equal allele sampling.
• To enable combined analysis of structural and nucleotide variation within a unified statistical framework.

Main Methods:

• Developed svgem, an expectation-maximization algorithm.
• Estimates allele and genotype frequencies and posterior probabilities.
• Tests for Hardy-Weinberg equilibrium and population differences using observed allele counts.

Main Results:

• svgem handles bi-allelic structural variation of any type detected by split reads or paired ends.
• The method accommodates arbitrary allele sampling bias.
• Validated with simulated data and real data from the 1000 Genomes Project.

Conclusions:

• svgem facilitates population distribution studies of structural variants using low-coverage sequencing data without prior genotyping.
• Enables integrated analysis of structural and nucleotide variation, avoiding genotype imputation biases.
• Advances the population genetics of structural variants by overcoming previous methodological limitations.